# Hailo > Gain valuable insights into AI technology and industry trends with Hailo's blog, a resource for tech enthusiasts and professionals alike. --- ## Posts - [Automatic License Plate Recognition with Hailo Processors](https://hailo.ai/blog/automatic-license-plate-recognition-with-hailo-processors/): Hailo's Automatic License Plate Recognition application is an end-to-end solution for deploying AI in intelligent transportation on the edge. - [Automatic License Plate Recognition with Hailo-8](https://hailo.ai/de/blog/automatic-license-plate-recognition-with-hailo-8/): Die automatische Nummernschilderkennung von Hailo ist eine End-to-End-Lösung für den Einsatz von KI im intelligenten Transportwesen am Rande. - [Automatic License Plate Recognition with Hailo Processors](https://hailo.ai/zh-hans/blog/automatic-license-plate-recognition-with-hailo-processors/): In this blog post, we present Hailo’s License Plate Recognition (LPR) implementation (also known as Automatic Number Place Recognition or... - [Automatic License Plate Recognition with Hailo Processors](https://hailo.ai/ja/blog/automatic-license-plate-recognition-with-hailo-processors/): In this blog post, we present Hailo’s License Plate Recognition (LPR) implementation (also known as Automatic Number Place Recognition or... - [AI Video Enhancement: Implications on Forensic Validity](https://hailo.ai/blog/ai-video-enhancement-implications-on-forensic-validity/): AI in forensic science & investigations offers speed but brings risks to integrity. Explore how to balance both in video evidence processing. - [AI Video Enhancement: Implications on Forensic Validity](https://hailo.ai/zh-hans/blog/ai-video-enhancement-implications-on-forensic-validity/): Artificial intelligence (AI) has revolutionized the way cameras process and enhance images, raising concerns about forensic authenticity. As AI-driven enhancements... - [AI Video Enhancement: Implications on Forensic Validity](https://hailo.ai/de/blog/ai-video-enhancement-implications-on-forensic-validity/): Artificial intelligence (AI) has revolutionized the way cameras process and enhance images, raising concerns about forensic authenticity. As AI-driven enhancements... - [AI Video Enhancement: Implications on Forensic Validity](https://hailo.ai/ja/blog/ai-video-enhancement-implications-on-forensic-validity/): Artificial intelligence (AI) has revolutionized the way cameras process and enhance images, raising concerns about forensic authenticity. As AI-driven enhancements... - [Hailo Hackathon 2024-2025: Pushing the Limits of AI Innovation on Raspberry Pi](https://hailo.ai/blog/hailo-hackathon-2024-2025-pushing-the-limits-of-ai-innovation-on-raspberry-pi/): Get inspired with the best hackathon ideas for AI! See how Hailo’s AI Hackathon teams built innovative projects using Raspberry Pi & Edge AI. - [Hailo Hackathon 2024-2025: Pushing the Limits of AI Innovation on Raspberry Pi](https://hailo.ai/de/blog/hailo-hackathon-2024-2025-pushing-the-limits-of-ai-innovation-on-raspberry-pi/): The third annual Hailo Hackathon was bigger, bolder, and more innovative than ever! Over 24 hours, 60 Hailo employees came... - [Hailo Hackathon 2024-2025: Pushing the Limits of AI Innovation on Raspberry Pi](https://hailo.ai/ja/blog/hailo-hackathon-2024-2025-pushing-the-limits-of-ai-innovation-on-raspberry-pi/): The third annual Hailo Hackathon was bigger, bolder, and more innovative than ever! Over 24 hours, 60 Hailo employees came... - [Hailo Hackathon 2024-2025: Pushing the Limits of AI Innovation on Raspberry Pi](https://hailo.ai/zh-hans/blog/hailo-hackathon-2024-2025-pushing-the-limits-of-ai-innovation-on-raspberry-pi/): The third annual Hailo Hackathon was bigger, bolder, and more innovative than ever! Over 24 hours, 60 Hailo employees came... - [CES 2025: The Year Edge AI Took Over](https://hailo.ai/blog/ces-2025-the-year-edge-ai-took-over/): From zero-code robotics to advanced surveillance, see how Hailo’s Edge AI demos dazzled CES 2025. Explore the latest AI trends & highlights! - [CES 2025: The Year Edge AI Took Over](https://hailo.ai/zh-hans/blog/ces-2025-the-year-edge-ai-took-over/): Walking through the halls of CES 2025, it was impossible to ignore the dominance of AI in nearly every consumer... - [CES 2025: The Year Edge AI Took Over](https://hailo.ai/de/blog/ces-2025-the-year-edge-ai-took-over/): Walking through the halls of CES 2025, it was impossible to ignore the dominance of AI in nearly every consumer... - [CES 2025: The Year Edge AI Took Over](https://hailo.ai/ja/blog/ces-2025-the-year-edge-ai-took-over/): Walking through the halls of CES 2025, it was impossible to ignore the dominance of AI in nearly every consumer... - [Scaling Up Video Management Systems  ](https://hailo.ai/blog/ai-enhanced-video-management-scalability/): Elevate your security with Hailo's scalable AI-driven video management systems for unparalleled surveillance performance. - [Scaling Up Video Management Systems  ](https://hailo.ai/de/blog/ai-enhanced-video-management-scalability/): What is a video management system? Video Management Systems, also known as VMS, collect inputs from multiple cameras and other... - [Scaling Up Video Management Systems  ](https://hailo.ai/ja/blog/ai-enhanced-video-management-scalability/): Hailo のスケーラブルな AI 駆動ビデオ管理システムでセキュリティを強化し、比類のない監視パフォーマンスを実現します。 - [Scaling Up Video Management Systems  ](https://hailo.ai/zh-hans/blog/ai-enhanced-video-management-scalability/): What is a video management system? Video Management Systems, also known as VMS, collect inputs from multiple cameras and other... - [Balancing Personal Privacy and Public Safety with Edge AI ](https://hailo.ai/blog/ai-in-public-safety-privacy-and-security/): Explore how AI in public safety balances security & privacy. Our edge AI solutions ensure secure environments while respecting privacy. Learn how! - [利用边缘人工智能平衡个人隐私与公共安全](https://hailo.ai/zh-hans/blog/ai-in-public-safety-privacy-and-security/): 探索公共安全领域的人工智能如何平衡安全性和隐私性。我们的边缘人工智能解决方案在确保安全环境的同时尊重隐私。了解如何做到！ - [Ausgleich zwischen persönlichem Datenschutz und öffentlicher Sicherheit mit Edge AI ](https://hailo.ai/de/blog/ai-in-public-safety-privacy-and-security/): Entdecken Sie, wie KI in öffentlichen Sicherheitslösungen sichere Umgebungen unter Wahrung der Privatsphäre bietet. Erfahren Sie wie! - [個人のプライバシーとエッジAIによる公共の安全のバランス](https://hailo.ai/ja/blog/ai-in-public-safety-privacy-and-security/): 公共の安全における AI がセキュリティとプライバシーのバランスをどのように保っているかをご覧ください。当社のエッジ AI ソリューションは、プライバシーを尊重しながら安全な環境を確保します。その方法をご覧ください。 - [The Evolution of AI on the Edge: From Perception to Creation](https://hailo.ai/blog/the-evolution-of-ai-on-the-edge-from-perception-to-creation/): Innovating at the technological forefront with Generative AI at the Edge. Learn about its role in the evolution of modern edge computing. - [The Evolution of AI on the Edge: From Perception to Creation](https://hailo.ai/zh-hans/blog/the-evolution-of-ai-on-the-edge-from-perception-to-creation/): In the vast landscape of artificial intelligence (AI), one of the most intriguing journeys has been the evolution of AI... - [The Evolution of AI on the Edge: From Perception to Creation](https://hailo.ai/de/blog/the-evolution-of-ai-on-the-edge-from-perception-to-creation/): In the vast landscape of artificial intelligence (AI), one of the most intriguing journeys has been the evolution of AI... - [The Evolution of AI on the Edge: From Perception to Creation](https://hailo.ai/ja/blog/the-evolution-of-ai-on-the-edge-from-perception-to-creation/): In the vast landscape of artificial intelligence (AI), one of the most intriguing journeys has been the evolution of AI... - [AI Smart Cameras: From Vision to Insights](https://hailo.ai/blog/ai-cameras-from-vision-to-insights/): The future of smart cameras with Hailo technologies combines AI algorithms with real-time smart video analytics for various applications. - [AI Smart Cameras: From Vision to Insights](https://hailo.ai/de/blog/ai-cameras-from-vision-to-insights/): As the camera market is booming, so does the need to empower cameras with artificial intelligence (AI). In this blog... - [AI Smart Cameras: From Vision to Insights](https://hailo.ai/zh-hans/blog/ai-cameras-from-vision-to-insights/): As the camera market is booming, so does the need to empower cameras with artificial intelligence (AI). In this blog... - [AI Smart Cameras: From Vision to Insights](https://hailo.ai/ja/blog/ai-cameras-from-vision-to-insights/): As the camera market is booming, so does the need to empower cameras with artificial intelligence (AI). In this blog... - [Backing into the Future: Unlocking the Potential of Automated Parking](https://hailo.ai/blog/backing-into-the-future-unlocking-the-potential-of-automated-parking/): Hailo AI’s autonomous parking solutions are here. Discover the futuristic parking lot technologies of automated parking AI. Learn more today! - [Backing into the Future: Unlocking the Potential of Automated Parking](https://hailo.ai/de/blog/backing-into-the-future-unlocking-the-potential-of-automated-parking/): The technology advancements and market drivers that accelerate the transition to automated parking - [Backing into the Future: Unlocking the Potential of Automated Parking](https://hailo.ai/zh-hans/blog/backing-into-the-future-unlocking-the-potential-of-automated-parking/): 由先进人工智能和传感器提供支持的自动停车正在改变我们的停车方式，减少事故并提高道路安全。 - [Backing into the Future: Unlocking the Potential of Automated Parking](https://hailo.ai/ja/blog/backing-into-the-future-unlocking-the-potential-of-automated-parking/): The technology advancements and market drivers that accelerate the transition to automated parking - [Leveraging Vendor Partnerships for ADAS Success: LeddarTech and Hailo](https://hailo.ai/blog/leveraging-vendor-partnerships-for-adas-success-leddartech-and-hailo/): Harnessing the power of ADAS development to deliver state-of-the-art AI driving solutions. Discover how we are paving the way for safer and smarter roads. - [Leveraging Vendor Partnerships for ADAS Success: LeddarTech and Hailo](https://hailo.ai/zh-hans/blog/leveraging-vendor-partnerships-for-adas-success-leddartech-and-hailo/): It’s a late summer evening, you’ve had a long day at work and all you want to do is get... - [Leveraging Vendor Partnerships for ADAS Success: LeddarTech and Hailo](https://hailo.ai/de/blog/leveraging-vendor-partnerships-for-adas-success-leddartech-and-hailo/): Nutzen Sie die Leistungsfähigkeit der ADAS-Entwicklung, um hochmoderne KI-Fahrlösungen bereitzustellen. Für sicherere und intelligentere Straßen. - [Leveraging Vendor Partnerships for ADAS Success: LeddarTech and Hailo](https://hailo.ai/ja/blog/leveraging-vendor-partnerships-for-adas-success-leddartech-and-hailo/): It’s a late summer evening, you’ve had a long day at work and all you want to do is get... - [AI Object Detection on the Edge: Making the Right Choice](https://hailo.ai/blog/ai-object-detection-on-the-edge-making-the-right-choice/): Advanced AI object detection for edge devices with Hailo. Learn how to make the right choices for your projects. Discover AI solutions today! - [AI Object Detection on the Edge: Making the Right Choice](https://hailo.ai/de/blog/object-detection-at-the-edge-making-the-right-choice/): Objekterkennungs-KI am Edge: Alles, was Sie wissen müssen, wenn Sie ein Objekterkennungsnetzwerk für Ihre Edge-Anwendung auswählen. - [AI Object Detection on the Edge: Making the Right Choice](https://hailo.ai/zh-hans/blog/ai-object-detection-on-the-edge-making-the-right-choice/): When choosing an AI object detection network for edge devices, there are many factors you should consider: compute power, memory resources, and... - [AI Object Detection on the Edge: Making the Right Choice](https://hailo.ai/ja/blog/ai-object-detection-on-the-edge-making-the-right-choice/): When choosing an AI object detection network for edge devices, there are many factors you should consider: compute power, memory resources, and... - [Multi-Camera Multi-Person Re-Identification with Hailo-8 ](https://hailo.ai/blog/multi-camera-multi-person-re-identification/): See how multi-camera tracking & multi-person tracking with Hailo-8 boosts operations for retail and security efficient accurate insights. - [Wiedererkennung mehrerer Personen mit mehreren Kameras mit Hailo-8 ](https://hailo.ai/de/blog/multi-camera-multi-person-re-identification/): Mehrkamera-Tracking und Mehrpersonen-Tracking mit Hailo-8 verbessern den Einzelhandelsbetrieb und liefern präzise Einblicke in die Sicherheit. - [使用Hailo-8进行多摄像头多人重识别 ](https://hailo.ai/zh-hans/blog/multi-camera-multi-person-re-identification/): 了解 Hailo-8 的多摄像机跟踪和多人跟踪如何促进零售和安全运营的高效准确洞察。 - [Hailo-8によるマルチカメラ複数人物再同定](https://hailo.ai/ja/blog/multi-camera-multi-person-re-identification/): Hailo-8 によるマルチカメラ追跡と複数人物追跡により、小売業やセキュリティ業務が効率化され、正確な洞察が得られる仕組みをご覧ください。 - [AI Video Analytics: A Cost-effective Edge Device for Small Businesses – an Achievable Challenge](https://hailo.ai/blog/advanced-video-analytics-for-small-businesses/): AI video analytics are a cost-effective solution enabling real-time accurate event detection for small businesses and smart retail. - [AI Video Analytics: A Cost-effective Edge Device for Small Businesses – an Achievable Challenge](https://hailo.ai/de/blog/advanced-video-analytics-for-small-businesses/): KI-Videoanalysen sind eine kostengünstige Lösung, die eine Ereigniserkennung in Echtzeit für den intelligenten Einzelhandel ermöglicht. - [AI Video Analytics: A Cost-effective Edge Device for Small Businesses – an Achievable Challenge](https://hailo.ai/zh-hans/blog/advanced-video-analytics-for-small-businesses/): Video Management Software (VMS) solutions are available in the market for more than a decade and some surveillance vendors are... - [AI Video Analytics: A Cost-effective Edge Device for Small Businesses – an Achievable Challenge](https://hailo.ai/ja/blog/advanced-video-analytics-for-small-businesses/): Video Management Software (VMS) solutions are available in the market for more than a decade and some surveillance vendors are... - [Powerful Edge AI to Boost Intelligent Traffic Monitoring](https://hailo.ai/blog/powerful-edge-ai-to-boost-intelligent-traffic-systems-its/): Hailo leads the way in traffic monitoring AI, transforming ITS into smarter, safer networks. Learn how in our detailed blog post. - [Powerful Edge AI to Boost Intelligent Traffic Monitoring](https://hailo.ai/de/blog/powerful-edge-ai-to-boost-intelligent-traffic-systems-its/): Whether in the city or outside of it, current traffic monitoring solutions have significant challenges in dealing with increasingly common... - [Powerful Edge AI to Boost Intelligent Traffic Monitoring](https://hailo.ai/zh-hans/blog/powerful-edge-ai-to-boost-intelligent-traffic-systems-its/): Whether in the city or outside of it, current traffic monitoring solutions have significant challenges in dealing with increasingly common... - [Powerful Edge AI to Boost Intelligent Traffic Monitoring](https://hailo.ai/ja/blog/powerful-edge-ai-to-boost-intelligent-traffic-systems-its/): Whether in the city or outside of it, current traffic monitoring solutions have significant challenges in dealing with increasingly common... - [How Software Can Streamline Edge AI Developer Experience](https://hailo.ai/blog/how-software-can-streamline-customer-experience-in-edge-ai/): Uncover the challenges of mass AI adoption, the role of vendor collaboration, and Hailo's innovative onboarding solutions. - [How Software Can Streamline Edge AI Developer Experience](https://hailo.ai/de/blog/how-software-can-streamline-customer-experience-in-edge-ai/): Entdecken Sie die Herausforderungen der Masseneinführung von KI, die Rolle der Zusammenarbeit mit Anbietern und die innovativen von Hailo. - [How Software Can Streamline Edge AI Developer Experience](https://hailo.ai/zh-hans/blog/how-software-can-streamline-customer-experience-in-edge-ai/): When looking at the challenges facing mass AI adoption, infrastructure and integration are among their most important aspects. They were... - [How Software Can Streamline Edge AI Developer Experience](https://hailo.ai/ja/blog/how-software-can-streamline-customer-experience-in-edge-ai/): When looking at the challenges facing mass AI adoption, infrastructure and integration are among their most important aspects. They were... - [5 Questions on the Edge (AI) with Hailo CBO Hadar Zeitlin](https://hailo.ai/blog/5-questions-on-the-edge-ai-with-hailo-cbo-hadar-zeitlin/): Join Hailo's CBO in exploring the edge AI landscape, its challenges, and the company's role in shaping its future. - [5 Questions on the Edge (AI) with Hailo CBO Hadar Zeitlin](https://hailo.ai/de/blog/5-questions-on-the-edge-ai-with-hailo-cbo-hadar-zeitlin/): Entdecken Sie gemeinsam mit dem CBO von Hailo die Edge-KI und die Rolle des Unternehmens bei der Gestaltung seiner Zukunft. - [5 Questions on the Edge (AI) with Hailo CBO Hadar Zeitlin](https://hailo.ai/zh-hans/blog/5-questions-on-the-edge-ai-with-hailo-cbo-hadar-zeitlin/): It is very exciting to see the increasing adoption of AI in edge applications in recent years and how the... - [5 Questions on the Edge (AI) with Hailo CBO Hadar Zeitlin](https://hailo.ai/ja/blog/5-questions-on-the-edge-ai-with-hailo-cbo-hadar-zeitlin/): Join Hailo's CBO in exploring the edge AI landscape, its challenges, and the company's role in shaping its future. - [Intelligent Video Analytics: A New Generation of Video Analytics Enabled by Powerful Edge AI](https://hailo.ai/blog/a-new-generation-of-video-analytics-enabled-by-powerful-edge-ai/): Intelligent Video Analytics: Learn how edge AI is becoming mainstream, providing real-time video analytics on intelligent cameras, NVRs, and edge AI boxes. - [Intelligent Video Analytics: A New Generation of Video Analytics Enabled by Powerful Edge AI](https://hailo.ai/de/blog/a-new-generation-of-video-analytics-enabled-by-powerful-edge-ai/): Intelligente Videoanalyse: Erfahren Sie, wie Edge AI Echtzeit-Videoanalysen für intelligente Kameras, NVRs und Edge AI-Boxen bereitstellt. - [Intelligent Video Analytics: A New Generation of Video Analytics Enabled by Powerful Edge AI](https://hailo.ai/zh-hans/blog/a-new-generation-of-video-analytics-enabled-by-powerful-edge-ai/): Intelligent Video Analytics (IVA) have penetrated many applications across industry verticals. Starting from on-premises server-based capabilities, then becoming available on... - [Intelligent Video Analytics: A New Generation of Video Analytics Enabled by Powerful Edge AI](https://hailo.ai/ja/blog/a-new-generation-of-video-analytics-enabled-by-powerful-edge-ai/): Intelligent Video Analytics (IVA) have penetrated many applications across industry verticals. Starting from on-premises server-based capabilities, then becoming available on... - [Play to Win: How Hailo and NXP Create Efficient Embedded AI Solutions](https://hailo.ai/blog/play-to-win-how-hailo-and-nxp-create-efficient-embedded-ai-solutions/): Learn about Hailo and NXP's partnership in developing efficient embedded AI solutions. Explore our innovative technologies. Read more now! - [Play to Win: How Hailo and NXP Create Efficient Embedded AI Solutions](https://hailo.ai/de/blog/play-to-win-how-hailo-and-nxp-create-efficient-embedded-ai-solutions/): Visit Hailo's blog to learn how integrating the powerful Hailo-8 AI processor into a range of NXP platforms has created some amazing computing solutions across applications and industries. - [Play to Win: How Hailo and NXP Create Efficient Embedded AI Solutions](https://hailo.ai/zh-hans/blog/play-to-win-how-hailo-and-nxp-create-efficient-embedded-ai-solutions/): A good partnership is all about the added value. It highlights the strengths of each party, compensates for the weaker... - [Play to Win: How Hailo and NXP Create Efficient Embedded AI Solutions](https://hailo.ai/ja/blog/play-to-win-how-hailo-and-nxp-create-efficient-embedded-ai-solutions/): A good partnership is all about the added value. It highlights the strengths of each party, compensates for the weaker... - [Pairing Sensing with AI for Efficient ADAS](https://hailo.ai/blog/pairing-sensing-with-ai-for-efficient-adas/): Hailo is revolutionizing ADAS sensing by integrating AI for enhanced efficiency. From sensor capabilities to real-world applications. - [Pairing Sensing with AI for Efficient ADAS](https://hailo.ai/de/blog/pairing-sensing-with-ai-for-efficient-adas/): Hailo revolutioniert die ADAS-Sensorik durch die Integration von KI für mehr Effizienz. Von Sensorfunktionen bis hin zu realen Anwendungen. - [Pairing Sensing with AI for Efficient ADAS](https://hailo.ai/zh-hans/blog/pairing-sensing-with-ai-for-efficient-adas/): Sensor and sensing capabilities are common and expanding in modern vehicles. One of the major motivators for this is safety... - [Pairing Sensing with AI for Efficient ADAS](https://hailo.ai/ja/blog/pairing-sensing-with-ai-for-efficient-adas/): Hailo is revolutionizing ADAS sensing by integrating AI for enhanced efficiency. From sensor capabilities to real-world applications. - [5 Questions on the Edge (AI) with Hailo CEO Orr Danon](https://hailo.ai/blog/5-questions-on-the-edge-ai-with-hailo-ceo-orr-danon/): Join Hailo CEO Orr Danon as he addresses five pivotal questions on Edge AI's impact, with insights on automotive autonomy, security, & more. - [5 Questions on the Edge (AI) with Hailo CEO Orr Danon](https://hailo.ai/de/blog/5-questions-on-the-edge-ai-with-hailo-ceo-orr-danon/): Begleiten Sie Orr Danon, CEO von Hailo, während er fünf zentrale Fragen zum Einfluss von Edge AI beantwortet. - [5 Questions on the Edge (AI) with Hailo CEO Orr Danon](https://hailo.ai/zh-hans/blog/5-questions-on-the-edge-ai-with-hailo-ceo-orr-danon/): Where do you see high-performance Edge AI making the most impact? Automotive autonomy and driver assistance (ADAS) immediately comes to... - [5 Questions on the Edge (AI) with Hailo CEO Orr Danon](https://hailo.ai/ja/blog/5-questions-on-the-edge-ai-with-hailo-ceo-orr-danon/): Where do you see high-performance Edge AI making the most impact? Automotive autonomy and driver assistance (ADAS) immediately comes to... - [Customer Case Study: Developing a High-Performance Application on an Embedded Edge AI Device](https://hailo.ai/blog/customer-case-study-developing-a-high-performance-application-on-an-embedded-edge-ai-device/): This AI case study explores the shift from AI/ML environment to real world app deployment on embedded edge AI device - [Customer Case Study: Developing a High-Performance Application on an Embedded Edge AI Device](https://hailo.ai/de/blog/customer-case-study-developing-a-high-performance-application-on-an-embedded-edge-ai-device/): Diese KI-Fallstudie untersucht den Wandel von der KI/ML-Umgebung zur realen App-Bereitstellung auf eingebetteten Edge-KI-Geräten. - [Customer Case Study: Developing a High-Performance Application on an Embedded Edge AI Device](https://hailo.ai/zh-hans/blog/customer-case-study-developing-a-high-performance-application-on-an-embedded-edge-ai-device/): One of the challenges in building embedded AI applications is taking it from machine learning research environment all the way... - [Customer Case Study: Developing a High-Performance Application on an Embedded Edge AI Device](https://hailo.ai/ja/blog/customer-case-study-developing-a-high-performance-application-on-an-embedded-edge-ai-device/): One of the challenges in building embedded AI applications is taking it from machine learning research environment all the way... - [Edge ML Deep Dive: Why You Should Use Tiles in Squeeze-and-Excite Operations](https://hailo.ai/blog/edge-ml-deep-dive-why-you-should-use-tiles-in-squeeze-and-excite-operations/): Edge machine learning is changing the game. Dive into our blog to understand how tiles optimize squeeze and excite operations for efficiency. - [Edge ML Deep Dive: Why You Should Use Tiles in Squeeze-and-Excite Operations](https://hailo.ai/de/blog/edge-ml-deep-dive-why-you-should-use-tiles-in-squeeze-and-excite-operations/): This blog post presents the Tiled Squeeze-and-Excite (TSE) – a method designed to improve the deployment efficiency of the Squeeze-and-Excite... - [Edge ML Deep Dive: Why You Should Use Tiles in Squeeze-and-Excite Operations](https://hailo.ai/zh-hans/blog/edge-ml-deep-dive-why-you-should-use-tiles-in-squeeze-and-excite-operations/): This blog post presents the Tiled Squeeze-and-Excite (TSE) – a method designed to improve the deployment efficiency of the Squeeze-and-Excite... - [Edge ML Deep Dive: Why You Should Use Tiles in Squeeze-and-Excite Operations](https://hailo.ai/ja/blog/edge-ml-deep-dive-why-you-should-use-tiles-in-squeeze-and-excite-operations/): This blog post presents the Tiled Squeeze-and-Excite (TSE) – a method designed to improve the deployment efficiency of the Squeeze-and-Excite... - [How an Open Model Zoo Can Boost Your Edge AI System Development](https://hailo.ai/blog/how-an-open-model-zoo-can-boost-your-edge-ai-system-development/): Empower your Edge AI development with Hailo's comprehensive overview of AI model zoos, including TensorFlow and ONNX ecosystems. - [Wie ein Open Model Zoo die Entwicklung Ihres Edge-KI-Systems vorantreiben kann](https://hailo.ai/de/blog/how-an-open-model-zoo-can-boost-your-edge-ai-system-development/): Stärken Sie Ihre Edge-KI-Entwicklung mit Hailos umfassendem Überblick über KI-Modellzoos, einschließlich TensorFlow- und ONNX-Ökosystemen. - [开放式Model Zoo如何促进边缘AI系统开发](https://hailo.ai/zh-hans/blog/how-an-open-model-zoo-can-boost-your-edge-ai-system-development/): 利用 Hailo 对 AI 模型库（包括 TensorFlow 和 ONNX 生态系统）的全面概述来增强您的边缘 AI 开发。 - [オープンModel ZooがエッジAIシステム開発を促進する方法](https://hailo.ai/ja/blog/how-an-open-model-zoo-can-boost-your-edge-ai-system-development/): TensorFlow や ONNX エコシステムを含む、Hailo の AI モデル ズーの包括的な概要を活用して、エッジ AI 開発を強化します。 - [Smart Retail on the Edge: Why Powerful Intelligent Vision is the Future of Brick & Mortar Stores](https://hailo.ai/blog/the-edge-of-retail-why-powerful-intelligent-vision-is-the-future-of-brick-mortar-stores/): See how smart retail solutions blend technology and retail to optimize shopping experiences. AI vision redefines the customer experience. - [Smart Retail on the Edge: Why Powerful Intelligent Vision is the Future of Brick & Mortar Stores](https://hailo.ai/de/blog/the-edge-of-retail-why-powerful-intelligent-vision-is-the-future-of-brick-mortar-stores/): Erfahren Sie, wie intelligente Einzelhandelslösungen Technologie und Einzelhandel vereinen, um das Einkaufserlebnis zu optimieren. - [Smart Retail on the Edge: Why Powerful Intelligent Vision is the Future of Brick & Mortar Stores](https://hailo.ai/zh-hans/blog/the-edge-of-retail-why-powerful-intelligent-vision-is-the-future-of-brick-mortar-stores/): Brick-and-mortar retail has been facing tough competition from eCommerce, with the COVID-19 tailwind reinforcing the growing transition from brick-and-mortar to... - [Smart Retail on the Edge: Why Powerful Intelligent Vision is the Future of Brick & Mortar Stores](https://hailo.ai/ja/blog/the-edge-of-retail-why-powerful-intelligent-vision-is-the-future-of-brick-mortar-stores/): Brick-and-mortar retail has been facing tough competition from eCommerce, with the COVID-19 tailwind reinforcing the growing transition from brick-and-mortar to... - [Mind the System Gap: System-Level Implications for High-Performance Edge AI Coprocessors](https://hailo.ai/blog/mind-the-system-gap-system-level-implications-for-high-performance-edge-ai-coprocessors/): Hailo delves into the implications of deploying edge AI co-processors, offering insights into achieving high-performance AI at the edge. - [Mind the System Gap: System-Level Implications for High-Performance Edge AI Coprocessors](https://hailo.ai/de/blog/mind-the-system-gap-system-level-implications-for-high-performance-edge-ai-coprocessors/): Visit Hailo's blog to learn about validation and system challenges, as the AI processor needs to deliver the best performance within a system’s constraints and without burdening its resources. - [Mind the System Gap: System-Level Implications for High-Performance Edge AI Coprocessors](https://hailo.ai/zh-hans/blog/mind-the-system-gap-system-level-implications-for-high-performance-edge-ai-coprocessors/): Intro The new generation of domain-specific AI computing architectures is booming. The need for these and the benefit they provide... - [Mind the System Gap: System-Level Implications for High-Performance Edge AI Coprocessors](https://hailo.ai/ja/blog/mind-the-system-gap-system-level-implications-for-high-performance-edge-ai-coprocessors/): Intro The new generation of domain-specific AI computing architectures is booming. The need for these and the benefit they provide... - [Evaluating Edge AI Accelerator Performance: Why TOPS Are Not Enough](https://hailo.ai/blog/evaluating-edge-ai-accelerator-performance-why-tops-are-not-enough/): Hailo sheds light on the significance of TOPS AI and why it's crucial to look beyond this metric when assessing AI accelerator performance. --- ## Resources - [Edge AI Processors for Smarter Security Systems – Exclusive Webinar with Hailo](https://hailo.ai/resources/industries/security/edge-ai-processors-for-smarter-security-systems-exclusive-webinar-with-hailo/): Watch Hailo’s recorded webinar available on-demand on AI video analytics and smart security system cameras. Real demos & expert insight. - [Hailo-8R Mini PCIe AI Acceleration Module Product Brief](https://hailo.ai/files/hailo-8r-mpcie-et-product-brief-en/): The Hailo-8 Mini PCIe Module is an AI accelerator module for AI applications, compatible with PCI Express Mini (mPCIe) form factor. - [Hailo-8 Century High Performance PCIe Cards](https://hailo.ai/files/hailo-8-century-pcie-product-brief-en/): Hailo has developed the best performing edge AI processor – a revolutionary deep learning solution that allows smart edge devices to run high-performance AI applications that could previously only run in the cloud. - [Hailo-8L Product Brief](https://hailo.ai/files/hailo-8l-product-brief-en/): Hailo has developed the best performing edge AI processor – a revolutionary deep learning solution that allows smart edge devices to run high-performance AI applications that could previously only run in the cloud. - [Hailo-8L M.2 Extended Temperature Product Brief](https://hailo.ai/files/hailo-8l-m-2-et-product-brief-en/): Hailo has developed the best performing edge AI processor – a revolutionary deep learning solution that allows smart edge devices to run high-performance AI applications that could previously only run in the cloud. - [TEMPO: AI-Generated Music Based on Heart Rate](https://hailo.ai/resources/industries/other/tempo-ai-generated-music-based-on-heart-rate/): Let your heart set the tempo! TEMPO uses AI to generate music based on your heartbeat. Experience TEMPO, Hailo’s real-time sound innovation. - [TAILO: AI-Powered Smart Pet Companion](https://hailo.ai/resources/industries/other/tailo-ai-powered-smart-pet-companion/): Meet TAILO, the smart AI pet companion using Hailo AI and Raspberry Pi. Track activity, reward behavior, and monitor your pet in real time. - [AD GENIE: Personalized Advertisement](https://hailo.ai/resources/industries/retail/ad-genie-personalized-advertisement/): Match your customer’s style to relevant ads in real time. Ad Genie uses CLIP model inference on Pi to create engaging shopping experiences. - [Ebook: Powerful Edge AI to Empower Intelligent Transportation System](https://hailo.ai/resources/industries/security/ebook-powerful-edge-ai-to-empower-intelligent-transportation-system/): Smarter transportation starts with Edge AI solutions! Explore AI-powered ITS in our free eBook & enhance traffic management. Download now! - [Luxonis License Plate Recognition for Smart Parking Solutions](https://hailo.ai/resources/industries/security/luxonis-license-plate-recognition-for-smart-parking-solutions/): Optimize parking management with AI-driven license plate recognition by Luxonis & Hailo AI. Click to watch our smart parking solution demo! - [Dell 3200 with AI empowered by Hailo](https://hailo.ai/resources/industries/security/dell-3200-with-ai-empowered-by-hailo/): AI-Enhanced Dell 3200 Gateway with Hailo brings advanced people detection and segmentation to rugged industrial settings. Watch the demo now! - [Hailo-10H demo - Stable Diffusion on the edge](https://hailo.ai/resources/industries/personal-compute/hailo-10h-demo-stable-diffusion-on-the-edge/): Watch offline generative AI in action with Hailo-10H accelerator. Run offline stable diffusion on the edge for fast, private image generation. - [Raspberry Pi AI Kit - Object recognition](https://hailo.ai/resources/industries/personal-compute/raspberry-pi-ai-kit-object-recognition/): Transform Raspberry Pi into a low-latency AI vision system. Use the HAT+ with Hailo-8L for real-time object recognition. Start building today! - [Raspberry Pi AI Kit - Pose estimation](https://hailo.ai/resources/industries/security/raspberry-pi-ai-kit-pose-estimation/): Raspberry Pi with Hailo accelerators enhance real-time pose estimation. Compact, power-efficient, & perfect for edge AI computer vision tasks. - [Raspberry Pi AI Kit - Object recognition](https://hailo.ai/resources/industries/security/raspberry-pi-ai-kit-object-recognition-2/): Enhance your Raspberry Pi with AI vision and real-time object detection. See how our kit brings low-latency AI to edge devices to developers. - [Raspberry Pi AI Kit - Subject segmentation](https://hailo.ai/resources/industries/automotive/raspberry-pi-ai-kit-subject-segmentation/): Experience real-time subject segmentation on the Raspberry Pi AI Kit powered by Hailo-8. Discover advanced edge AI—watch the demo now! - [System Electronics Astrial Board on a Drone platform](https://hailo.ai/resources/industries/security/system-electronics-astrial-board-on-a-drone-platform/): Transform surveillance with advanced drone face detection. Powered by Hailo-8 AI and featuring real-time analytics—click to see the demo! - [PercivAI Radar-Based 3D Perception](https://hailo.ai/resources/industries/automotive/percivai-radar-based-3d-perception/): Enhance vehicle safety with advanced radar 3D perception. Explore our innovative automotive demo featuring PercivAI and Hailo AI solutions. - [Tier IV: Lidar and camera sensor fusion for high accuracy perception](https://hailo.ai/resources/industries/automotive/tier-iv-lidar-and-camera-sensor-fusion-for-high-accuracy-perception/): Discover next-gen sensor fusion with Tier IV & Hailo AI's lidar and camera fusion demo. High accuracy and on-device anonymization! Watch now! - [TT Control ruggedized ECU for off-road vehicles](https://hailo.ai/resources/industries/automotive/tt-control-ruggedized-ecu-for-off-road-vehicles/): Enhance worker safety with our camera-based ECU for off-road vehicles and industrial machines. Explore Hailo's AI innovative tech today! - [Limelight 4: AI-Powered Robot Controller](https://hailo.ai/resources/industries/security/limelight-4-ai-powered-robot-controller/): Limelight 4 makes AI accessible with its zero-code smart camera and robot controller, powered by Hailo-8 & Raspberry Pi presented at CES 2025. - [Hailo Company Brochure - Portuguese](https://hailo.ai/files/hailo-company-brochure-pt/): Hailo has developed the best performing edge AI processor – a revolutionary deep learning solution that allows smart edge devices to run high-performance AI applications that could previously only run in the cloud. - [Hailo Company Brochure](https://hailo.ai/files/hailo-company-brochure-en/): Hailo has developed the best performing edge AI processor – a revolutionary deep learning solution that allows smart edge devices to run high-performance AI applications that could previously only run in the cloud. - [Happy Holidays from Hailo](https://hailo.ai/resources/industries/other/happy-holidays-from-hailo/): Happy Holidays from Hailo AI! View our special AI-powered holiday video greeting inspired by the CLIP model. 🎁 Watch now and share the joy! - [Hailo Company Brochure - English](https://hailo.ai/files/hailo-company-brochure-en/): Hailo has developed the best performing edge AI processor – a revolutionary deep learning solution that allows smart edge devices to run high-performance AI applications that could previously only run in the cloud. - [THRIVE LOGIC: Hockey Arena Security System](https://hailo.ai/resources/industries/security/thrive-logic-hockey-arena-security-system/): Discover how Hailo AI's security solution improved hockey arena safety with enhanced AI video analytics and monitoring. Read the case study. - [THRIVE: 冰球场馆安保系统](https://hailo.ai/zh-hans/resources/industries/security-zh-hans/thrive-logic-hockey-arena-security-system/): 了解 Hailo AI 的安全解决方案如何通过增强的 AI 视频分析和监控功能提升冰球场馆的安全性。阅读案例研究。Thrive Logic 软件支持视频管理和分析运行应用程序。 - [THRIVE: Sicherheitssystem für Hockey-Arena](https://hailo.ai/de/resources/industries/security-de/thrive-hockey-arena-security-system/): Entdecken Sie, wie die Sicherheitslösung von Hailo AI durch verbesserte KI-Videoanalyse die Sicherheit in Eishockeyarenen verbessert hat. - [THRIVE: ホッケーアリーナセキュリティシステム](https://hailo.ai/ja/resources/industries/security-ja/thrive-hockey-arena-security-system/): Hailo AI のセキュリティ ソリューションが、強化された AI ビデオ分析と監視によってホッケー アリーナの安全性をどのように向上させたかをご覧ください。ケース スタディをお読みください。 - [AutoSens Interview with Yaniv Sulkes, VP of Automotive](https://hailo.ai/resources/industries/automotive/autosens-interview-with-yaniv-sulkes-vp-of-automotive/): Learn how AI is driving automotive change with Hailo’s VP, Yaniv Sulkes, in this AutoSens interview. Discover the future of mobility now. - [VMS with 100+ channels](https://hailo.ai/resources/industries/security/vms-with-100-channels/): Improve your video management system (VMS) with Hailo's AI processors, handling up to 100 channels with low latency and power efficiency. - [The Rise and Future Trends of Edge Computing and AI](https://hailo.ai/resources/industries/security/the-rise-and-future-trends-of-edge-computing-and-ai/): Explore future trends of AI in edge computing for security, medical, production quality control, and more. Watch our Hailo AI expert webinar! - [The Rise and Future Trends of Edge Computing and AI](https://hailo.ai/zh-hans/resources/industries/security-zh-hans/the-rise-and-future-trends-of-edge-computing-and-ai/): 探索边缘计算领域人工智能在安全、医疗、生产质量控制等领域的未来趋势。观看我们的 Hailo AI 专家网络研讨会！ - [The Rise and Future Trends of Edge Computing and AI](https://hailo.ai/de/resources/industries/security-de/the-rise-and-future-trends-of-edge-computing-and-ai/): Entdecken Sie zukünftige Trends der KI im Edge Computing für Sicherheit, Medizin, Produktionsqualitätskontrolle und mehr. Jetzt ansehen! - [The Rise and Future Trends of Edge Computing and AI](https://hailo.ai/ja/resources/industries/security-ja/the-rise-and-future-trends-of-edge-computing-and-ai/): セキュリティ、医療、生産品質管理などのエッジコンピューティングにおける AI の将来のトレンドを探ります。Hailo AI エキスパート ウェビナーをご覧ください。 - [Velo AI's Copilot Elevates Bike Safety with AI](https://hailo.ai/resources/industries/automotive/velo-ais-copilot-elevates-bike-safety-with-ai/): Explore how Velo AI's CoPilot uses AI technology to improve bike safety. Elevate cycling with smart bike solutions. Learn more today! - [Velo AI's Copilot Elevates Bike Safety with AI](https://hailo.ai/zh-hans/resources/industries/automotive-zh-hans/velo-ais-copilot-elevates-bike-safety-with-ai/): 探索 Velo AI 的 CoPilot 如何利用 AI 技术提高自行车安全性。利用智能自行车解决方案提升骑行体验。立即了解更多信息！ - [Velo AI's Copilot Elevates Bike Safety with AI](https://hailo.ai/de/resources/industries/automotive-de/velo-ais-copilot-elevates-bike-safety-with-ai/): Entdecken Sie, wie Velo AIs CoPilot und Hailo AI KI-Technologie mit intelligenten Fahrradlösungen nutzen. Erfahren Sie noch heute mehr! - [Velo AI's Copilot Elevates Bike Safety with AI](https://hailo.ai/ja/resources/industries/automotive-ja/velo-ais-copilot-elevates-bike-safety-with-ai/): Velo AI の CoPilot が AI テクノロジーを活用して自転車の安全性を向上させる方法をご覧ください。スマート バイク ソリューションでサイクリングを向上させましょう。今すぐ詳細をご確認ください。 - [Hailo-10 AI Accelerators for the Generative AI Era](https://hailo.ai/resources/industries/personal-compute/hailo-10-ai-accelerators-for-the-generative-ai-era/): Video demo on Hailo-10 accelerated Generative AI application examples for content creation. Boost PC productivity & unleash creativity! - [Hailo-10 AI Accelerators for the Generative AI Era](https://hailo.ai/de/resources/industries/personal-compute-de/hailo-10-ai-accelerators-for-the-generative-ai-era/): Videodemo zu Hailo-10 beschleunigten generative-KI Anwendungsbeispiele für die Inhaltserstellung. Steigern Sie die PC-Produktivität! - [Designing powerful, scalable & cost-efficient AI-powered video management systems](https://hailo.ai/resources/industries/security/designing-powerful-scalable-cost-efficient-ai-powered-video-management-systems/): Discover the power of video analytics in industrial security with our detailed eBook. A must-read for industry innovators & technologists. - [Designing powerful, scalable & cost-efficient AI-powered video management systems](https://hailo.ai/de/resources/industries/security-de/designing-powerful-scalable-cost-efficient-ai-powered-video-management-systems/): Entdecken Sie die Leistungsfähigkeit der Videoanalyse in der industriellen Sicherheit mit unserem eBook für Innovatoren und Technologen. - [VMS with CLIP-based free text search](https://hailo.ai/resources/industries/security/vms-with-clip-based-free-text-search/): Discover next generation VMS with Hailo & Network Optix. Zero shot free text search on live video streams. Learn more on the future of VMS! - [VMS with CLIP-based free text search](https://hailo.ai/zh-hans/resources/industries/security-zh-hans/vms-with-clip-based-free-text-search/): 使用 Hailo 和 Network Optix 探索下一代 VMS。实时视频流上的零样本免费文本搜索。详细了解 VMS 的未来！ - [VMS with CLIP-based free text search](https://hailo.ai/de/resources/industries/security-de/vms-with-clip-based-free-text-search/): Entdecken Sie die nächste VMS-Generation mit Hailo & Network Optix. Zero-Shot-Freitextsuche in Live-Videostreams. Erfahren Sie mehr! - [VMS with CLIP-based free text search](https://hailo.ai/ja/resources/industries/security-ja/vms-with-clip-based-free-text-search/): Hailo と Network Optix を使用した 次世代 VMS をご覧ください。ライブ ビデオ ストリームでのゼロ ショットの無料テキスト検索。VMS の将来についてさらに詳しく学びましょう。 - [Leonardo Elsag Automatic License Plate Reader for Law Enforcement](https://hailo.ai/resources/industries/security/leonardo-elsag-automatic-license-plate-reader-for-law-enforcement/): Improve public safety with Leonardo ELSAG Automatic License Plate Reader. Explore our advanced ALPR solutions for law enforcement. Visit now! - [Leonardo Elsag Automatic License Plate Reader for Law Enforcement](https://hailo.ai/zh-hans/resources/industries/security-zh-hans/leonardo-elsag-automatic-license-plate-reader-for-law-enforcement/): 使用 Leonardo ELSAG 自动车牌读取器提高公共安全。探索我们为执法部门提供的先进 ALPR 解决方案。立即访问！ - [Leonardo Elsag Automatic License Plate Reader for Law Enforcement](https://hailo.ai/de/resources/industries/security-de/leonardo-elsag-automatic-license-plate-reader-for-law-enforcement/): Verbessern Sie die öffentliche Sicherheit mit der Lösung Automatischer Kennzeichenleser (ALPR) für die Strafverfolgung. Jetzt besuchen! - [Leonardo Elsag Automatic License Plate Reader for Law Enforcement](https://hailo.ai/ja/resources/industries/security-ja/leonardo-elsag-automatic-license-plate-reader-for-law-enforcement/): Leonardo ELSAG 自動ナンバープレート リーダーで公共の安全を向上しましょう。法執行機関向けの高度な ALPR ソリューションをご覧ください。今すぐアクセスしてください。 - [B&R AI Smart Camera for Real Time Automation](https://hailo.ai/resources/industries/industrial-automation/ai-smart-camera-for-real-time-automation/): Harness the power of smart cameras with Hailo AI for real-time automation security and surveillance, simple to integrate and simple to apply. - [Gunsens Shooting Threat Detector](https://hailo.ai/resources/industries/security/gunsens-shooting-threat-detector/): Enhance your security measures with AI shooting threat detectors powered by Hailo, designed to provide unparalleled safety in public spaces. - [B&R AI Smart Camera for Real Time Automation](https://hailo.ai/zh-hans/resources/industries/industrial-automation-zh-hans/ai-smart-camera-for-real-time-automation/): 利用 Hailo AI 智能摄像头的强大功能实现实时自动化安全和监控，易于集成且易于应用。 - [B&R AI Smart Camera for Real Time Automation](https://hailo.ai/de/resources/industries/industrial-automation-de/ai-smart-camera-for-real-time-automation/): Nutzen Sie die Leistung intelligenter Kameras mit Hailo AI für automatisierte Sicherheit und Überwachung in Echtzeit, einfach zu integrieren. - [B&R AI Smart Camera for Real Time Automation](https://hailo.ai/ja/resources/industries/industrial-automation-ja/ai-smart-camera-for-real-time-automation/): Hailo AI を搭載したスマート カメラのパワーを活用して、リアルタイムの自動化セキュリティと監視を実現し、簡単に統合して簡単に適用できます。 - [Unleashing the Power of AI Edge Processing](https://hailo.ai/resources/industries/security/unleashing-the-power-of-ai-edge-processing/): Discover the revolutionary impact of AI-powered edge processing, from improved real-time analysis to heightened system reliability. - [Unleashing the Power of AI Edge Processing](https://hailo.ai/de/resources/industries/security-de/unleashing-the-power-of-ai-edge-processing/): Entdecken Sie die Auswirkungen der KI-gestützten Edge-Verarbeitung, von verbesserten Echtzeitanalysen bis hin zur Systemzuverlässigkeit. - [Hailo Quality Declaration](https://hailo.ai/wp-content/uploads/2024/02/Hailo-Quality-Policy.pdf): Hailo AI stands for exceptional quality in AI technology, crafting reliable processors that power innovation and drive the industry forward. - [Bird's Eye View 3D Perception Solution](https://hailo.ai/resources/industries/automotive/birds-eye-view/): Discover how Hailo’s Bird's-Eye View technology improves automotive safety and driver assistance, offering 360-degree camera applications. - [3D-Wahrnehmungslösung aus der Vogelperspektive](https://hailo.ai/de/resources/industries/automotive-de/birds-eye-view-3d-perception-solution/): Entdecken Sie, wie die Vogelperspektive-Technologie von Hailo die Fahrzeugsicherheit verbessert und 360-Grad-Kameraanwendungen bietet. - [Bird's Eye View 3D Perception Solution](https://hailo.ai/ja/resources/industries/automotive-ja/birds-eye-view/): Hailo の Bird's-Eye View テクノロジーが 360 度カメラ アプリケーションを提供し、自動車の安全性と運転支援をどのように向上させるかをご覧ください。 - [Bird’s Eye View 3D Perception Solution](https://hailo.ai/zh-hans/resources/industries/automotive-zh-hans/birds-eye-view-3d-perception-solution/): 了解 Hailo 的鸟瞰技术如何通过提供 360 度摄像头应用来提高汽车安全和驾驶员辅助。 - [L2+/L3 Full Surround Perception for ADAS and Automated Driving](https://hailo.ai/resources/industries/automotive/l2-l3-full-surround-perception-for-adas-and-automated-driving/): Hailo 8 provides real-time processing with high accuracy & low latency for L2-L3 automated driving. Start driving safer today with AI! - [L2+/L3 Vollständige Surround-Wahrnehmung für ADAS und automatisiertes Fahren](https://hailo.ai/de/resources/industries/automotive-de/l2-l3-full-surround-perception-for-adas-and-automated-driving/): Hailos Full Surround Perception bietet eine umfassende Abdeckung für autonomes Fahren L2-L3. Fahren Sie noch heute sicherer! - [L2+/L3 Full Surround Perception for ADAS and Automated Driving](https://hailo.ai/ja/resources/industries/automotive-ja/l2-l3-full-surround-perception-for-adas-and-automated-driving/): Hailo のフル サラウンド パーセプションは、L2 ～ L3 自動運転を包括的にカバーします。 AI を活用して今すぐ安全運転を始めましょう! - [L2+/L3 Full Surround Perception for ADAS and Automated Driving](https://hailo.ai/zh-hans/resources/industries/automotive-zh-hans/l2-l3-full-surround-perception-for-adas-and-automated-driving/): Hailo的全环绕感知为L2-L3自动驾驶提供全面覆盖。 今天就开始利用人工智能更安全地驾驶吧！ - [AI-Powered Low-Light Video Enhancement](https://hailo.ai/resources/industries/security/ai-powered-low-light-video-enhancement/): How to enhance low light video using AI that brings a revolution in video enhancement, turning low light scenes into bright, clear visuals. - [AI-Powered Low-Light Video Enhancement](https://hailo.ai/ja/resources/industries/security-ja/ai-powered-low-light-video-enhancement/): AI を使用して低照度ビデオを強化し、ビデオ強化に革命をもたらし、低照度シーンを明るく鮮明なビジュアルに変える方法。 - [AI-Powered Low-Light Video Enhancement](https://hailo.ai/zh-hans/resources/industries/security-zh-hans/ai-powered-low-light-video-enhancement/): 如何使用人工智能增强低光视频，从而带来视频增强的革命，将低光场景变成明亮、清晰的视觉效果。 - [KI-gestützte Videoverbesserung bei schwachen Lichtverhältnissen](https://hailo.ai/de/resources/industries/security-de/ai-powered-low-light-video-enhancement/): Verbessern Sie Videos bei schlechten Lichtverhältnissen mit KI, die eine Revolution in der Videoverbesserung mit sich bringt. - [Intelligent Ultrasound: NeedleTrainer](https://hailo.ai/resources/industries/other/intelligent-ultrasound-needletrainer/): The customer story of Intelligent Ultrasound & Hailo's collaboration on AI-driven innovation in medical imaging and ultrasound training. - [Intelligent Ultrasound: NeedleTrainer](https://hailo.ai/de/resources/industries/other-de/intelligent-ultrasound-needletrainer/): Die Zusammenarbeit von Intelligent Ultrasound und Hailo bei KI-gesteuerten Innovationen in der medizinischen Bildgebung und im Ultraschall. - [Intelligent Ultrasound: NeedleTrainer](https://hailo.ai/ja/resources/industries/other-ja/intelligent-ultrasound-needletrainer/): インテリジェント超音波と Hailo のコラボレーションに関する説得力のある顧客ストーリーを調査し、医療画像における AI 主導のイノベーションへの道を明らかにします。 ヘルスケア技術における私たちのパートナーシップの顕著な成果と革命的な影響を掘り下げてみましょう。 - [Intelligent Ultrasound: NeedleTrainer](https://hailo.ai/zh-hans/resources/industries/other-zh-hans/intelligent-ultrasound-needletrainer/): 探索智能超声与 Hailo 合作的引人入胜的客户故事，阐明人工智能驱动的医学成像创新之路。 深入研究我们的合作伙伴关系对医疗保健技术的卓越成就和革命性影响。 - [Hailo-15 Vision Processors – Unprecedented AI Performance in a Camera Power Envelope](https://hailo.ai/resources/industries/industrial-automation/hailo-15-vision-processors-unprecedented-ai-performance-in-a-camera-power-envelope/): Discover Hailo-15 Vision Processors, delivering unparalleled AI performance for cameras, optimizing efficiency and automation capabilities. - [Hailo's solutions for High Performance Video Management System (VMS)](https://hailo.ai/resources/industries/security/hailos-solutions-for-high-performance-video-management-system-vms/): Discover how Hailo is revolutionizing the security industry with state-of-the-art VMS solutions, innovations, features, and benefits. - [Hailos Lösungen für High Performance Video Management System (VMS)](https://hailo.ai/de/resources/industries/security-de/hailos-solutions-for-high-performance-video-management-system-vms/): Entdecken Sie, wie Hailo die Sicherheitsbranche mit hochmodernen VMS-Lösungen, Innovationen, Funktionen und Vorteilen revolutioniert. - [Hailo-15 Vision Processors – Unprecedented AI Performance in a Camera Power Envelope](https://hailo.ai/de/resources/industries/security-de/hailo-15-vision-processors-unprecedented-ai-performance-in-a-camera-power-envelope/): Entdecken Sie Hailo-15 Vision-Prozessoren, die beispiellose KI-Leistung für Kameras bieten und die Effizienz optimieren. - [Scalable & powerful platform for premium L2+ and higher ADAS ECUs](https://hailo.ai/resources/industries/automotive/scalable-powerful-platform-for-premium-l2-and-higher-adas-ecus/): ADAS ECUs with Hailo's scalable and powerful platform. Tailored for L2 and higher systems, ensuring superior performance and safety. - [Scalable & powerful platform for premium L2+ and higher ADAS ECUs](https://hailo.ai/de/resources/industries/automotive-de/scalable-powerful-platform-for-premium-l2-and-higher-adas-ecus/): ADAS-Steuergeräte mit der skalierbaren und leistungsstarken Plattform von Hailo. Für L2- und höhere Systeme zur Gewährleistung der Sicherheit. - [Ebook: Powerful Video Analytics at Scale](https://hailo.ai/resources/industries/industrial-automation/ebook-powerful-video-analytics-at-scale/): Advanced video management system solutions for high-security environments. Edge analytics for faster detection and actionable insights - [Art of Logic: City+ Smart City Security Solution](https://hailo.ai/resources/industries/security/art-of-logic-city-smart-city-security-solution/): Hailo AI is at the forefront of smart city security, delivering AI-powered solutions for enhanced safety in public and private urban spaces. - [Art of Logic: City+ Smart City Security Solution](https://hailo.ai/de/resources/industries/security-de/art-of-logic-city-smart-city-security-solution/): Hailo AI steht an der Spitze der Smart-City-Sicherheit und liefert KI-gestützte Lösungen für mehr Sicherheit im städtischen Raum. - [Art of Logic: City+ Smart City Security Solution](https://hailo.ai/ja/resources/industries/security-ja/art-of-logic-city-smart-city-security-solution/): Hailo AI はスマート シティ セキュリティの最前線にあり、公共および私有の都市空間の安全性を強化する AI を活用したソリューションを提供しています。 - [Art of Logic: City+ Smart City Security Solution](https://hailo.ai/zh-hans/resources/industries/security-zh-hans/art-of-logic-city-smart-city-security-solution/): Hailo AI 处于智慧城市安全的前沿，提供人工智能驱动的解决方案，以增强公共和私人城市空间的安全。 - [Demonstration of the Hailo-15 Family of High-performance AI Vision Processors](https://youtu.be/TxWPoWvd_k4?si=9wdkyIG-cLldFxsl): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Demonstration of the Hailo-15 Family of High-performance AI Vision Processors](https://youtu.be/TxWPoWvd_k4?si=9wdkyIG-cLldFxsl): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Demonstration of the Hailo-15 Family of High-performance AI Vision Processors](https://youtu.be/TxWPoWvd_k4?si=9wdkyIG-cLldFxsl): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Demonstration of the Hailo-15 Family of High-performance AI Vision Processors](https://youtu.be/TxWPoWvd_k4?si=9wdkyIG-cLldFxsl): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Demonstration of Art of Logic’s Video Management System and the latest edge AI and vision technologies and products](https://youtu.be/jqrARrwj1k0): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Demonstration of Smart Cameras for Intelligent Transportation Systems](https://youtu.be/FNxdPmXnfe4): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Demonstration of Smart Cameras for Intelligent Transportation Systems](https://youtu.be/FNxdPmXnfe4): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Demonstration of Art of Logic’s Video Management System and the latest edge AI and vision technologies and products](https://youtu.be/jqrARrwj1k0): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Demonstration of Smart Cameras for Intelligent Transportation Systems](https://youtu.be/FNxdPmXnfe4): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Demonstration of Art of Logic’s Video Management System and the latest edge AI and vision technologies and products](https://youtu.be/jqrARrwj1k0): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Demonstration of Smart Cameras for Intelligent Transportation Systems](https://youtu.be/FNxdPmXnfe4): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Demonstration of Art of Logic’s Video Management System and the latest edge AI and vision technologies and products](https://youtu.be/jqrARrwj1k0): Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - [Idein: AI Cast Smart Camera](https://hailo.ai/resources/industries/retail/idein-ai-cast-smart-camera/): Discover how Hailo's Idein AI Cast smart camera is transforming the retail industry with advanced AI capabilities for retail analytics. - [Idein: AI Cast Smart Camera](https://hailo.ai/de/resources/industries/security-de/idein-ai-cast-smart-camera/): Die intelligente Kamera „Idein AI Cast“ von Hailo verändert den Einzelhandel mit KI-Funktionen für Einzelhandelsanalysen. - [Idein /AISIN 採用事例](https://hailo.ai/ja/resources/industries/security-ja/idein-ai-cast-smart-camera/): Hailo の Idein AI Cast スマート カメラが、小売分析用の高度な AI 機能によって小売業界をどのように変革しているかをご覧ください。 --- ## Products - [Model Explorer](https://hailo.ai/products/hailo-software/model-explorer/): A dynamic tool designed to help users explore the models on Hailo's Model Zoo and select the best NN models for their AI applications - [Hailo-10H M.2 Generative AI Acceleration Module](https://hailo.ai/products/ai-accelerators/hailo-10h-m-2-generative-ai-acceleration-module/): The generative AI accelerator Hailo-10H M.2 module improves AI inferencing on edge devices, ideal for PCs, smart cars, and edge computing. - [Hailo-10H M.2 Generative AI Acceleration Module](https://hailo.ai/zh-hans/products/ai-accelerators/hailo-10h-m-2-generative-ai-acceleration-module/): 生成式 AI 加速器改变边缘设备上的 AI 推理。 高效的 Hailo-10H M.2 非常适合 PC、智能汽车和边缘计算。 - [Hailo-10H M.2 Generative AI Acceleration Module](https://hailo.ai/de/products/ai-accelerators/hailo-10h-m-2-generative-ai-acceleration-module/): Der generative KI-Beschleuniger transformiert die KI-Inferenz auf Edge-Geräten. Ideal für PCs, Smart Cars und Edge Computing. - [Hailo-10H M.2 Generative AI Acceleration Module](https://hailo.ai/ja/products/ai-accelerators/hailo-10h-m-2-generative-ai-acceleration-module/): エッジ デバイスでの AI 推論を変革する生成 AI アクセラレータ。 効率的な Hailo-10H M.2 は、PC、スマートカー、エッジ コンピューティングに最適です. - [Quality and Reliability](https://hailo.ai/ja/products/quality-and-reliability/): 品質、耐久性、信頼性の基盤に基づいて構築された AI 製品による卓越性への Hailo AI の取り組みについて学びましょう。 - [Quality and Reliability](https://hailo.ai/de/products/quality-and-reliability/): Erfahren Sie mehr über die KI-Exzellenz von Hailo mit KI-Produkten, die auf Qualität und Zuverlässigkeit basieren. - [Quality and Reliability](https://hailo.ai/products/quality-and-reliability/): Learn about Hailo AI's commitment to excellence with AI products built on a foundation of quality, durability, and reliability. - [Quality and Reliability](https://hailo.ai/zh-hans/products/quality-and-reliability/): 了解 Hailo AI 对以质量、耐用性和可靠性为基础构建的 AI 产品的卓越承诺。 - [Software](https://hailo.ai/products/software/): Entdecken Sie die fortschrittlichen KI-Softwarelösungen von Hailo, die auf KI- und Bildverarbeitungsprozessoren zugeschnitten sind. - [Model Explorer - Computer Vision](https://hailo.ai/products/hailo-software/model-explorer-vision/): Discover Hailo’s AI Model Explorer for selecting the top NN models. Available in TensorFlow and ONNX formats. Explore our model zoo now! - [Hailo-8R mPCIe AI Acceleration module](https://hailo.ai/ja/products/ai-accelerators/hailo-8r-mpcie-ai-acceleration-module/): Hailo-8R Mini PCIe アクセラレータ モジュールは、PCI Express Mini (mPCIe) フォーム ファクタと互換性のある AI アプリケーション用の AI アクセラレータ モジュールです。 業界をリードする AI パフォーマンスをエッジ デバイスに提供 – 高い電力効率で最大 13 テラオペレーション/秒 (TOPS)。 - [Vision Processor Software Package](https://hailo.ai/products/hailo-software/vision-processor-software-package/): Hailo's Vision Processor Software, tailored for SoC, offers drivers, libraries, and tools for smart camera AI computer vision development. - [Hailo-8L Entry-Level AI Accelerator](https://hailo.ai/zh-hans/products/ai-accelerators/hailo-8l-ai-accelerator-for-ai-light-applications/): Hailo-8L 入门级人工智能加速器支持高达 13 TOPS、低延迟和功效。 简单的硬件集成解决方案。 - [Hailo-8L Entry-Level AI Accelerator](https://hailo.ai/ja/products/ai-accelerators/hailo-8l-ai-accelerator-for-ai-light-applications/): Hailo-8L エントリーレベル AI アクセラレーターは、最大 13 TOPS、低遅延、電力効率をサポートします。 簡単なハードウェア統合ソリューション。 - [Hailo-8 M.2 AI Acceleration Module](https://hailo.ai/zh-hans/products/ai-accelerators/hailo-8-m2-ai-acceleration-module/): Hailo-8™ M.2 AI 模块是一款与 NGFF M.2 外形尺寸兼容的 AI 加速器。 基于具有高能效的 26 TOPS 处理器。 - [Hailo-8 M.2 AI Acceleration Module](https://hailo.ai/ja/products/ai-accelerators/hailo-8-m2-ai-acceleration-module/): Hailo-8 M.2 AI モジュールは、NGFF M.2 フォームファクターと互換性のある AI アクセラレータです。 高い電力効率を備えた 26 TOPS プロセッサをベースとしています。 - [Hailo-8L M.2 Entry-Level Acceleration Module](https://hailo.ai/zh-hans/products/ai-accelerators/hailo-8l-m-2-ai-acceleration-module-for-ai-light-applications/): Hailo-8L M.2 AI 模块为边缘设备提供入门级 AI 加速，提供 13 TOPS 性能以及功耗和成本效率。 - [Hailo-8L M.2 Entry-Level Acceleration Module](https://hailo.ai/ja/products/ai-accelerators/hailo-8l-m-2-ai-acceleration-module-for-ai-light-applications/): Hailo-8L M.2 AI モジュールは、エッジ デバイスにエントリーレベルの AI アクセラレーションを提供し、電力とコスト効率に優れた 13 TOPS のパフォーマンスを提供します。 - [Hailo-8R mPCIe AI Acceleration module](https://hailo.ai/zh-hans/products/ai-accelerators/hailo-8r-mpcie-ai-acceleration-module/): Hailo-8R Mini PCIe 加速器模块是一款适用于 AI 应用的 AI 加速器模块，兼容 PCI Express Mini (mPCIe) 外形尺寸。 - [Hailo-8L Entry-Level AI Accelerator](https://hailo.ai/products/ai-accelerators/hailo-8l-ai-accelerator-for-ai-light-applications/): Hailo-8L Entry-Level AI Accelerator supports up to 13 TOPS, low latency, & power efficiency. Easy hardware integration solutions. - [Hailo-8 Century High Performance PCIe Card](https://hailo.ai/zh-hans/products/ai-accelerators/hailo-8-century-high-performance-pcie-card/): Hailo-8 Century AI加速卡提供高达208 TOPS的AI性能，具有最佳功率的实时深度神经网络推理。 - [Hailo-8 Century High Performance PCIe Card](https://hailo.ai/ja/products/ai-accelerators/hailo-8-century-high-performance-pcie-card/): Hailo-8 Century AI アクセラレータ カードは、最大 208 TOPS の AI パフォーマンス、最適なパワーでリアルタイムのディープ ニューラル ネットワーク推論を提供します。 - [Hailo-8 M.2 AI Acceleration Module](https://hailo.ai/products/ai-accelerators/hailo-8-m2-ai-acceleration-module/): Hailo-8 M.2 AI Module is an AI accelerator compatible with NGFF M.2 form factor. Based on a 26 TOPS processor with high power efficiency. - [AI Accelerators](https://hailo.ai/products/ai-accelerators/): Unleash the full potential of your AI applications with the high-performance, energy-efficient AI Accelerators from Hailo. - [Vision Processor Software Package](https://hailo.ai/zh-hans/products/hailo-software/vision-processor-software-package/): Hailo 的视觉处理器软件专为 SoC 量身定制，为智能相机 AI 计算机视觉开发提供驱动程序、库和工具。 - [Vision Processor Software Package](https://hailo.ai/ja/products/hailo-software/vision-processor-software-package/): SoC 向けにカスタマイズされた Hailo のビジョン プロセッサ ソフトウェアは、スマート カメラ AI コンピューター ビジョン開発用のドライバー、ライブラリ、ツールを提供します。 - [AI Vision Processors](https://hailo.ai/products/ai-vision-processors/): Experience the power of Hailo 15's AI vision processors. Designed for the modern world, our processors offer advanced AI imaging solutions. - [Hailo-8 Century High Performance PCIe Card](https://hailo.ai/products/ai-accelerators/hailo-8-century-high-performance-pcie-card/): Hailo-8 Century AI accelerator cards offer up to 208 TOPS of AI performance, real-time deep neural network inferencing with optimal power. - [AI Vision Processors](https://hailo.ai/zh-hans/products/ai-vision-processors/): 体验 Hailo 15 AI 视觉处理器的强大功能。 我们的处理器专为现代世界而设计，提供先进的人工智能成像解决方案。 - [AI Vision Processors](https://hailo.ai/de/products/ai-vision-processors/): AI-Vision-Prozessoren des Hailo 15 wurden für die moderne Welt entwickelt und bieten fortschrittliche KI-Bildgebungs lösungen. - [AI Vision Processors](https://hailo.ai/ja/products/ai-vision-processors/): Hailo 15 の AI ビジョン プロセッサーのパワーを体験してください。 現代世界向けに設計された当社のプロセッサーは、高度な AI イメージング ソリューションを提供します。 - [AI Accelerators](https://hailo.ai/zh-hans/products/ai-accelerators/): 利用 Hailo 的高性能、高能效 AI 加速器释放 AI 应用的全部潜力。 - [AI Accelerators](https://hailo.ai/de/products/ai-accelerators/): Entfesseln Sie das volle Potenzial Ihrer KI-Anwendungen mit den leistungsstarken, energieeffizienten KI-Beschleunigern von Hailo. - [AI Accelerators](https://hailo.ai/ja/products/ai-accelerators/): の高性能でエネルギー効率の高い AI アクセラレータを使用して、AI アプリケーションの可能性を最大限に引き出します。製品に関するお問い合わせは、Hailo AI の専門家にお問い合わせください。 - [Hailo-8R mPCIe AI Acceleration module](https://hailo.ai/products/ai-accelerators/hailo-8r-mpcie-ai-acceleration-module/): The Hailo-8R Mini PCIe Accelerator Module is an AI accelerator module for AI applications, compatible with PCI Express Mini (mPCIe) form factor. Delivering industry-leading AI performance for edge devices – up to 13 tera-operations per second (TOPS) with high power efficiency. - [Hailo-8L Entry-Level AI Accelerator](https://hailo.ai/de/products/ai-accelerators/hailo-8l-entry-level-ai-accelerator/): Der Hailo-8L™ KI-Beschleuniger zeichnet sich durch eine außergewöhnlich niedrige Latenz und hocheffiziente Verarbeitung aus. - [Hailo-8 M.2 AI Acceleration Module](https://hailo.ai/de/products/ai-accelerators/hailo-8-m-2-ai-acceleration-module/): Hailo-8 M.2 AI-Modul ist ein KI-Beschleuniger, der mit dem NGFF M.2-Formfaktor kompatibel ist. 26-TOPS-Prozessor mit hoher Energieeffizienz. - [Hailo-8L™ M.2 Entry-Level Acceleration Module](https://hailo.ai/de/products/ai-accelerators/hailo-8l-m-2-entry-level-acceleration-module/): Das M.2-KI-Beschleunigungsmodul verfügt über eine PCIe Gen 3.0 2-Lane-Schnittstelle für eine höchst begrenzte KI-Leistung für Edge-Geräte. - [Hailo-8R™ mPCIe AI Acceleration module](https://hailo.ai/de/products/ai-accelerators/hailo-8r-mpcie-ai-acceleration-module/): Das Hailo-8R Mini PCIe Accelerator Module ist ein KI-Beschleunigermodul für KI-Anwendungen, kompatibel mit dem PCI Express Mini (mPCIe)-Formfaktor. Bereitstellung branchenführender KI-Leistung für Edge-Geräte – bis zu 13 Tera-Operationen pro Sekunde (TOPS) bei hoher Energieeffizienz. - [Hailo-8 Century High Performance PCIe Card](https://hailo.ai/de/products/ai-accelerators/hailo-8-century-high-performance-pcie-card/): Die KI-Beschleunigerkarte Hailo-8 Century bietet bis zu 208 TOPS an KI-Leistung, Echtzeit-Deep-Neuronale-Netzwerk-Inferenz mit optimaler Leistung. - [Hailo-8L M.2 Entry-Level Acceleration Module](https://hailo.ai/products/ai-accelerators/hailo-8l-m-2-ai-acceleration-module-for-ai-light-applications/): The Hailo-8L M.2 AI Module delivers entry-level AI acceleration for edge devices, offering 13 TOPS performance with power & cost efficiency. - [Hailo AI Vision Processor Software Package](https://hailo.ai/de/products/hailo-software/hailo-ai-vision-processor-software-package/): Die auf SoC zugeschnittene Vision Processor Software von Hailo bietet Treiber, Bibliotheken & Tools für die KI-Computer-Vision-Entwicklung intelligenter Kameras. - [Hailo-8 AI Accelerator](https://hailo.ai/zh-hans/products/ai-accelerators/hailo-8-ai-accelerator/): Hailo-8 AI 处理器拥有 26 TOPS 硬件，在机器学习方面超越了其他 GPU。 在一分钱大小的空间内具有无与伦比的效率。 - [Software](https://hailo.ai/ja/products/hailo-software/): AI およびビジョンプロセッサ向けにカスタマイズされ、高い精度とリアルタイムの結果を保証する Hailo の高度な AI ソフトウェア ソリューションをご確認ください。 - [Software](https://hailo.ai/products/hailo-software/): Explore Hailo's advanced AI software solutions, tailored for AI & vision processors, ensuring high accuracy and real-time results. - [Hailo-8 AI Accelerator](https://hailo.ai/products/ai-accelerators/hailo-8-ai-accelerator/): The Hailo-8 AI processor, with 26 TOPS hardware, surpasses other GPUs for machine learning. With unmatched efficiency in a penny-sized space. - [Hailo-8 AI Accelerator](https://hailo.ai/ja/products/ai-accelerators/hailo-8-ai-accelerator/): 26 TOPS ハードウェアを備えた Hailo-8 AI プロセッサーは、機械学習において他の GPU を上回ります。 ペニーサイズのスペースで比類のない効率を実現します。 - [Hailo-8 AI Accelerator](https://hailo.ai/de/products/ai-accelerators/hailo-8-ai-accelerator/): Der Hailo-8 KI-Prozessor mit 26 TOPS-Hardware übertrifft andere GPUs beim maschinellen Lernen. Mit unübertroffener Effizienz auf kleinstem Raum. - [Software](https://hailo.ai/de/products/hailo-software/): Entdecken Sie die fortschrittlichen KI-Softwarelösungen von Hailo, die auf KI- und Bildverarbeitungsprozessoren zugeschnitten sind. - [Software](https://hailo.ai/zh-hans/products/hailo-software/): 探索 Hailo 先进的 AI 软件解决方案，专为 AI 和视觉处理器量身定制，确保高精度和实时结果。 - [Hailo-15 AI Vision Processor](https://hailo.ai/ja/products/ai-vision-processors/hailo-15-ai-vision-processor/): Hailo-15 AI ビジョン プロセッサは、スマート カメラ用に設計された複数の複雑な深層学習 AI アプリケーションを処理でき、最大 20 TOPS と比類のない AI ビデオ分析を提供します。 優れた画質と柔軟で安全なシステムを備えた Hailo-15 AI ビジョン プロセッサは、さまざまなカメラ向けのソリューションです。 最先端の ISP パイプラインと、ハイ ダイナミック レンジ (HDR) およびノイズ リダクション (NR) アルゴリズムを備えた高度なビジョン サブシステムにより、プレミアム 4K60 画質を提供します。 Hailo-15H、Hailo-15M、Hailo-15L の 3 つの製品からお選びください。 - [Hailo-15 AI Vision Processor](https://hailo.ai/zh-hans/products/ai-vision-processors/hailo-15-ai-vision-processor/): Hailo-15 AI 视觉处理器：智能相机最高可达 20 TOPS。 优质 4K60 画质、HDR、NR，提供三种型号 - Hailo-15H、M、L。 - [Hailo-15 AI Vision Processor](https://hailo.ai/de/products/ai-vision-processors/hailo-15-ai-vision-processor/): Der Hailo-15 KI Vision Processor kann mehrere komplexe Deep-Learning-KI-Anwendungen verarbeiten, die für intelligente Kameras entwickelt wurden, und bietet bis zu 20 TOPS und beispiellose KI-Videoanalysen. Mit überragender Bildqualität und einem flexiblen, sicheren System ist der Hailo-15 AI Vision Processor die Lösung für eine Vielzahl von Kameras. Bietet erstklassige 4K60-Bildqualität mit einer hochmodernen ISP-Pipeline und einem fortschrittlichen Vision-Subsystem mit Algorithmen für High Dynamic Range (HDR) und Rauschunterdrückung (NR). Wählen Sie aus drei Produktangeboten: Hailo-15H, Hailo-15M und Hailo-15L. - [Hailo-15 AI Vision Processor](https://hailo.ai/products/ai-vision-processors/hailo-15-ai-vision-processor/): Hailo-15 AI Vision Processor with up to 20 TOPS for smart cameras & edge applications. Learn more about the best VPUs for real-time imaging. - [Hailo AI Software Suite](https://hailo.ai/ja/products/hailo-software/hailo-ai-software-suite/): AI アプリケーションと深層学習モデルの導入用のソフトウェア。 ML フレームワークの統合と x86、ARM、Hailo AI プロセッサーの最適化。 - [Hailo AI Software Suite](https://hailo.ai/zh-hans/products/hailo-software/hailo-ai-software-suite/): 用于人工智能应用和深度学习模型部署的软件。 ML 框架集成和 x86、ARM、Hailo AI 处理器优化。 - [Hailo AI Software Suite](https://hailo.ai/de/products/hailo-software/hailo-ai-software-suite/): Software für KI-Anwendungen und die Bereitstellung von Deep-Learning-Modellen. Integration von ML-Frameworks & optimierte x86-, ARM- & Hailo-KI-Prozessoren - [Hailo AI Software Suite](https://hailo.ai/products/hailo-software/hailo-ai-software-suite/): Software for AI applications & deep learning model deployment. ML frameworks integration & x86, ARM, Hailo AI processors optimized. - [Technology](https://hailo.ai/ja/products/technology/): Hailo のニューラル ネットワーク プロセッサは、特殊なテクノロジー スタック設計を採用しており、ディープ ラーニング タスク用のフォン ノイマン アーキテクチャを最大限に実行します。 - [Technology](https://hailo.ai/zh-hans/products/technology/): Hailo 的神经网络处理器拥有专门的技术堆栈设计，在深度学习任务方面的性能优于冯诺依曼架构。 - [Technology](https://hailo.ai/products/technology/): Hailo’s neural network processor has a specialized technology stack design & out performs Von Neumann architecture for deep learning tasks. - [Technology](https://hailo.ai/de/products/technology/): Der neuronale Netzwerkprozessor von Hailo verfügt über ein spezielles Technologie-Stack-Design und übertrifft die Von-Neumann-Architektur für Deep-Learning-Aufgaben. --- # # Detailed Content ## Posts > Hailo's Automatic License Plate Recognition application is an end-to-end solution for deploying AI in intelligent transportation on the edge. - Published: 2025-04-09 - Modified: 2025-06-17 - URL: https://hailo.ai/blog/automatic-license-plate-recognition-with-hailo-processors/ - Categories: AI Software, Automatic License Plate Recognition, Edge AI Developer, Smart Transportation - Tags: LPR - Translation Priorities: Optional In this blog post, we present Hailo’s License Plate Recognition (LPR) implementation (also known as Automatic Number Place Recognition or ANPR). ALPR, or Automatic License Plate Recognition, is a technology that uses cameras and specialized software to automatically capture images of vehicle license plates and convert the alphanumeric characters into digital data. This data can then be instantly compared against various databases for a multitude of purposes.  In this blog post, we present Hailo’s Automatic License Plate Recognition (ALPR) implementation (also known as License / Number Plate Recognition or LPR / NPR). ALPR is a ubiquitous pipeline found in nearly all outdoor deployments. It is commonly used in two scenarios: integrated directly within the camera itself, or running on a ruggedized processing device connected to one or more cameras. This ALPR solution is ideal for Intelligent Transportation Systems (ITS) as well as law enforcement systems, and demonstrates how Hailo empowers real-life machine learning deployment in AI-based products. This blog post focuses on the second scenario, showcasing a complete, deployable AI pipeline based on Hailo-8.  The first scenario, where LPR runs directly on the camera, can be implemented with our Hailo-15 high-performance vision. The hardware configuration described here includes a full HD camera, a camera processor module, a Hailo-8 AI M. 2 module and a GStreamer application integrating the Computer Vision (CV) pipeline with multiple neural networks. Understanding Automatic License Plate Recognition Automatic License Plate Recognition (ALPR) system is one of the most popular video analytics applications for smart cities. Deployed on highways, toll booths, and parking lots, ALPR enables rapid vehicle identification, congestion control, vehicle counting, law enforcement control, automatic fare collection, and more.   Figure 1 -... --- > Die automatische Nummernschilderkennung von Hailo ist eine End-to-End-Lösung für den Einsatz von KI im intelligenten Transportwesen am Rande. - Published: 2025-04-09 - Modified: 2024-06-23 - URL: https://hailo.ai/de/blog/automatic-license-plate-recognition-with-hailo-8/ - Categories: AI Software, AI Software, Automatic License Plate Recognition, Edge AI Developer, Edge AI Developer, Smart Transportation - Tags: LPR - Translation Priorities: Optional In this blog post, we present Hailo’s License Plate Recognition (LPR) implementation (also known as Automatic Number Place Recognition or ANPR). In this blog post, we present Hailo’s License Plate Recognition (LPR) implementation (also known as Automatic Number Plate Recognition or ANPR). The presented solution can be used in Intelligent Transportation Systems (ITS) and is a good example of how Hailo-8 is being utilized in a real-life deployment of machine learning in AI-based products. We distinguish between two different deployment scenarios. One where the LPR pipeline runs on the camera and the other where there is a ruggedized processing device that is connected to one or more cameras that are feeding it. In this blog, we are focusing on the former case to highlight the possibility of enabling even the stringent constraints imposed by a camera-attached system unlocked by the capabilities of a high-performance AI processor. The device includes a full HD camera, camera processor, Hailo-8™ AI processor, and GStreamer application, which integrates Computer Vision (CV) pipeline with multi-neural networks. Introduction Automatic License Plate Recognition (ALPR) system is one of the most popular video analytics applications for smart cities. The system can be deployed on highways, toll booths, and parking lots to enable fast vehicle identification, congestion control, vehicle counting, law enforcement control, automatic fare collection, and more. Figure 1 - ALPR system output. The system is able to detect and track the vehicles as well as detect their license plates and recognize them With a powerful edge AI processor, ALPR can be deployed on edge devices and run in real-time, which is crucial for: Improving product miss-rates with better performing... --- - Published: 2025-04-09 - Modified: 2025-06-17 - URL: https://hailo.ai/zh-hans/blog/automatic-license-plate-recognition-with-hailo-processors/ - Categories: AI Software, Automatic License Plate Recognition, Edge AI Developer, Smart Transportation - Tags: LPR - Translation Priorities: 可选 In this blog post, we present Hailo’s License Plate Recognition (LPR) implementation (also known as Automatic Number Place Recognition or ANPR). ALPR, or Automatic License Plate Recognition, is a technology that uses cameras and specialized software to automatically capture images of vehicle license plates and convert the alphanumeric characters into digital data. This data can then be instantly compared against various databases for a multitude of purposes.  In this blog post, we present Hailo’s Automatic License Plate Recognition (ALPR) implementation (also known as License / Number Plate Recognition or LPR / NPR). ALPR is a ubiquitous pipeline found in nearly all outdoor deployments. It is commonly used in two scenarios: integrated directly within the camera itself, or running on a ruggedized processing device connected to one or more cameras. This ALPR solution is ideal for Intelligent Transportation Systems (ITS) as well as law enforcement systems, and demonstrates how Hailo empowers real-life machine learning deployment in AI-based products. This blog post focuses on the second scenario, showcasing a complete, deployable AI pipeline based on Hailo-8.  The first scenario, where LPR runs directly on the camera, can be implemented with our Hailo-15 high-performance vision. The hardware configuration described here includes a full HD camera, a camera processor module, a Hailo-8 AI M. 2 module and a GStreamer application integrating the Computer Vision (CV) pipeline with multiple neural networks. Understanding Automatic License Plate Recognition Automatic License Plate Recognition (ALPR) system is one of the most popular video analytics applications for smart cities. Deployed on highways, toll booths, and parking lots, ALPR enables rapid vehicle identification, congestion control, vehicle counting, law enforcement control, automatic fare collection, and more.   Figure 1 -... --- - Published: 2025-04-09 - Modified: 2025-06-17 - URL: https://hailo.ai/ja/blog/automatic-license-plate-recognition-with-hailo-processors/ - Categories: AI Software, Automatic License Plate Recognition, Edge AI Developer, Smart Transportation - Tags: LPR - Translation Priorities: 可选 In this blog post, we present Hailo’s License Plate Recognition (LPR) implementation (also known as Automatic Number Place Recognition or ANPR). ALPR, or Automatic License Plate Recognition, is a technology that uses cameras and specialized software to automatically capture images of vehicle license plates and convert the alphanumeric characters into digital data. This data can then be instantly compared against various databases for a multitude of purposes.  In this blog post, we present Hailo’s Automatic License Plate Recognition (ALPR) implementation (also known as License / Number Plate Recognition or LPR / NPR). ALPR is a ubiquitous pipeline found in nearly all outdoor deployments. It is commonly used in two scenarios: integrated directly within the camera itself, or running on a ruggedized processing device connected to one or more cameras. This ALPR solution is ideal for Intelligent Transportation Systems (ITS) as well as law enforcement systems, and demonstrates how Hailo empowers real-life machine learning deployment in AI-based products. This blog post focuses on the second scenario, showcasing a complete, deployable AI pipeline based on Hailo-8.  The first scenario, where LPR runs directly on the camera, can be implemented with our Hailo-15 high-performance vision. The hardware configuration described here includes a full HD camera, a camera processor module, a Hailo-8 AI M. 2 module and a GStreamer application integrating the Computer Vision (CV) pipeline with multiple neural networks. Understanding Automatic License Plate Recognition Automatic License Plate Recognition (ALPR) system is one of the most popular video analytics applications for smart cities. Deployed on highways, toll booths, and parking lots, ALPR enables rapid vehicle identification, congestion control, vehicle counting, law enforcement control, automatic fare collection, and more.   Figure 1 -... --- > AI in forensic science & investigations offers speed but brings risks to integrity. Explore how to balance both in video evidence processing. - Published: 2025-03-30 - Modified: 2025-06-17 - URL: https://hailo.ai/blog/ai-video-enhancement-implications-on-forensic-validity/ - Categories: Generative AI, Video Enhancement - Translation Priorities: Optional https://open. spotify. com/show/2Eiui4qdEizbmn6bx1Hyq3 Artificial intelligence (AI) has revolutionized the way cameras process and enhance images, raising concerns about forensic authenticity. As AI-driven enhancements become commonplace in forensic science, surveillance, and security applications, a critical question arises: does introducing AI-powered algorithms to video footage compromise its evidentiary integrity? The answer lies in understanding the distinction between the broad range of image enhancement functions and the algorithmic techniques applied to the source —from de-noising and sharpening to re-coloring, inpainting, and up to full image (and video) generation. AI and Forensically Authentic Images Recent advancements in AI have integrated intelligent image processing into the vision pipeline, improving video quality by reducing noise, enhancing sharpness, and refining details. However, this has triggered debate about whether such enhancements alter the original scene in a way that affects forensic validity. For an image or video to be considered forensically authentic, it must: Accurately represent the original scene without material alteration. Maintain a verifiable chain of custody from capture to presentation. Preserve intact metadata to indicate any modifications. Undergo technical validation to confirm that no unauthorized alterations impact its evidentiary significance. In court settings and forensic investigations, authenticity often relies on establishing: The photo's provenance (who took it, when, where, and with what equipment) That it has not been altered in ways that change its evidentiary significance That proper handling procedures were followed to maintain integrity Organizations like the Scientific Working Group on Digital Evidence (SWGDE) and National Institute of Standards and Technology (NIST) set guidelines for... --- - Published: 2025-03-30 - Modified: 2025-04-01 - URL: https://hailo.ai/zh-hans/blog/ai-video-enhancement-implications-on-forensic-validity/ - Categories: Generative AI, Video Enhancement - Translation Priorities: 可选 Artificial intelligence (AI) has revolutionized the way cameras process and enhance images, raising concerns about forensic authenticity. As AI-driven enhancements become commonplace in forensic science, surveillance, and security applications, a critical question arises: does introducing AI-powered algorithms to video footage compromise its evidentiary integrity? The answer lies in understanding the distinction between the broad range of image enhancement functions and the algorithmic techniques applied to the source —from de-noising and sharpening to re-coloring, inpainting, and up to full image (and video) generation. AI and Forensically Authentic Images Recent advancements in AI have integrated intelligent image processing into the vision pipeline, improving video quality by reducing noise, enhancing sharpness, and refining details. However, this has triggered debate about whether such enhancements alter the original scene in a way that affects forensic validity. For an image or video to be considered forensically authentic, it must: Accurately represent the original scene without material alteration. Maintain a verifiable chain of custody from capture to presentation. Preserve intact metadata to indicate any modifications. Undergo technical validation to confirm that no unauthorized alterations impact its evidentiary significance. In court settings and forensic investigations, authenticity often relies on establishing: The photo's provenance (who took it, when, where, and with what equipment) That it has not been altered in ways that change its evidentiary significance That proper handling procedures were followed to maintain integrity Organizations like the Scientific Working Group on Digital Evidence (SWGDE) and National Institute of Standards and Technology (NIST) set guidelines for handling digital photographic... --- - Published: 2025-03-30 - Modified: 2025-04-01 - URL: https://hailo.ai/de/blog/ai-video-enhancement-implications-on-forensic-validity/ - Categories: Generative AI, Video Enhancement - Translation Priorities: Optional Artificial intelligence (AI) has revolutionized the way cameras process and enhance images, raising concerns about forensic authenticity. As AI-driven enhancements become commonplace in forensic science, surveillance, and security applications, a critical question arises: does introducing AI-powered algorithms to video footage compromise its evidentiary integrity? The answer lies in understanding the distinction between the broad range of image enhancement functions and the algorithmic techniques applied to the source —from de-noising and sharpening to re-coloring, inpainting, and up to full image (and video) generation. AI and Forensically Authentic Images Recent advancements in AI have integrated intelligent image processing into the vision pipeline, improving video quality by reducing noise, enhancing sharpness, and refining details. However, this has triggered debate about whether such enhancements alter the original scene in a way that affects forensic validity. For an image or video to be considered forensically authentic, it must: Accurately represent the original scene without material alteration. Maintain a verifiable chain of custody from capture to presentation. Preserve intact metadata to indicate any modifications. Undergo technical validation to confirm that no unauthorized alterations impact its evidentiary significance. In court settings and forensic investigations, authenticity often relies on establishing: The photo's provenance (who took it, when, where, and with what equipment) That it has not been altered in ways that change its evidentiary significance That proper handling procedures were followed to maintain integrity Organizations like the Scientific Working Group on Digital Evidence (SWGDE) and National Institute of Standards and Technology (NIST) set guidelines for handling digital photographic... --- - Published: 2025-03-30 - Modified: 2025-04-01 - URL: https://hailo.ai/ja/blog/ai-video-enhancement-implications-on-forensic-validity/ - Categories: Generative AI, Video Enhancement - Translation Priorities: 可选 Artificial intelligence (AI) has revolutionized the way cameras process and enhance images, raising concerns about forensic authenticity. As AI-driven enhancements become commonplace in forensic science, surveillance, and security applications, a critical question arises: does introducing AI-powered algorithms to video footage compromise its evidentiary integrity? The answer lies in understanding the distinction between the broad range of image enhancement functions and the algorithmic techniques applied to the source —from de-noising and sharpening to re-coloring, inpainting, and up to full image (and video) generation. AI and Forensically Authentic Images Recent advancements in AI have integrated intelligent image processing into the vision pipeline, improving video quality by reducing noise, enhancing sharpness, and refining details. However, this has triggered debate about whether such enhancements alter the original scene in a way that affects forensic validity. For an image or video to be considered forensically authentic, it must: Accurately represent the original scene without material alteration. Maintain a verifiable chain of custody from capture to presentation. Preserve intact metadata to indicate any modifications. Undergo technical validation to confirm that no unauthorized alterations impact its evidentiary significance. In court settings and forensic investigations, authenticity often relies on establishing: The photo's provenance (who took it, when, where, and with what equipment) That it has not been altered in ways that change its evidentiary significance That proper handling procedures were followed to maintain integrity Organizations like the Scientific Working Group on Digital Evidence (SWGDE) and National Institute of Standards and Technology (NIST) set guidelines for handling digital photographic... --- > Get inspired with the best hackathon ideas for AI! See how Hailo’s AI Hackathon teams built innovative projects using Raspberry Pi & Edge AI. - Published: 2025-03-12 - Modified: 2025-06-17 - URL: https://hailo.ai/blog/hailo-hackathon-2024-2025-pushing-the-limits-of-ai-innovation-on-raspberry-pi/ - Categories: AI Hardware, Compute, Developer, Edge AI Device, Generative AI, Raspberry Pi 5 with Hailo AI HAT+ - Translation Priorities: Optional The third annual Hailo Hackathon was bigger, bolder, and more innovative than ever! Over 24 hours, 60 Hailo employees came together to push the boundaries of edge AI using the Hailo AI HAT+ (26TOPS) on the Raspberry Pi 5. This wasn’t just a coding event—it was a celebration of creativity, collaboration, and problem-solving. With great food, an overnight coding marathon, and an electric atmosphere of team spirit, this hackathon proved that AI innovation can be both exciting and impactful. Participants developed real-world AI applications that are now available for the community to explore and expand upon. You can find the projects in our Community Projects. Note: Some of the projects are built to work only with the 26 TOPS Hailo AI HAT+. Check out our Hackathon video How We Judged the Projects Our expert panel evaluated the projects based on several key criteria: Diversity & Creativity – Unique approaches and teamwork. Extra points were given to teams with members from various departments, such as marketing, design, and product management, showcasing the value of diverse perspectives and non-coding skills. Real-World Impact – Potential applications and benefits in everyday life. Implementation Quality – Code efficiency, performance, and robustness. Diversity & Creativity – Unique approaches and teamwork. Extra points were given to teams with members from various departments, such as marketing, design, and product management, showcasing the value of diverse perspectives and non-coding skills. Hackathon Takeaways The Hailo Hackathon 2024-2025 demonstrated that powerful AI applications can be developed in just 24 hours with the right tools. The Hailo AI HAT+ for Raspberry Pi provided teams... --- - Published: 2025-03-12 - Modified: 2025-06-17 - URL: https://hailo.ai/de/blog/hailo-hackathon-2024-2025-pushing-the-limits-of-ai-innovation-on-raspberry-pi/ - Categories: AI Hardware, Compute, Developer, Edge AI Device, Generative AI, Raspberry Pi 5 with Hailo AI HAT+ - Translation Priorities: Optional The third annual Hailo Hackathon was bigger, bolder, and more innovative than ever! Over 24 hours, 60 Hailo employees came together to push the boundaries of edge AI using the Hailo AI HAT+ (26TOPS) on the Raspberry Pi 5. This wasn’t just a coding event—it was a celebration of creativity, collaboration, and problem-solving. With great food, an overnight coding marathon, and an electric atmosphere of team spirit, this hackathon proved that AI innovation can be both exciting and impactful. Participants developed real-world AI applications that are now available for the community to explore and expand upon. You can find the projects in our Community Projects. Note: Some of the projects are built to work only with the 26 TOPS Hailo AI HAT+. Check out our Hackathon video How We Judged the Projects Our expert panel evaluated the projects based on several key criteria: Diversity & Creativity – Unique approaches and teamwork. Extra points were given to teams with members from various departments, such as marketing, design, and product management, showcasing the value of diverse perspectives and non-coding skills. Real-World Impact – Potential applications and benefits in everyday life. Implementation Quality – Code efficiency, performance, and robustness. Diversity & Creativity – Unique approaches and teamwork. Extra points were given to teams with members from various departments, such as marketing, design, and product management, showcasing the value of diverse perspectives and non-coding skills. Hackathon Takeaways The Hailo Hackathon 2024-2025 demonstrated that powerful AI applications can be developed in just 24 hours with the right tools. The Hailo AI HAT+ for Raspberry Pi provided teams... --- - Published: 2025-03-12 - Modified: 2025-06-17 - URL: https://hailo.ai/ja/blog/hailo-hackathon-2024-2025-pushing-the-limits-of-ai-innovation-on-raspberry-pi/ - Categories: AI Hardware, Compute, Developer, Edge AI Device, Generative AI, Raspberry Pi 5 with Hailo AI HAT+ - Translation Priorities: 可选 The third annual Hailo Hackathon was bigger, bolder, and more innovative than ever! Over 24 hours, 60 Hailo employees came together to push the boundaries of edge AI using the Hailo AI HAT+ (26TOPS) on the Raspberry Pi 5. This wasn’t just a coding event—it was a celebration of creativity, collaboration, and problem-solving. With great food, an overnight coding marathon, and an electric atmosphere of team spirit, this hackathon proved that AI innovation can be both exciting and impactful. Participants developed real-world AI applications that are now available for the community to explore and expand upon. You can find the projects in our Community Projects. Note: Some of the projects are built to work only with the 26 TOPS Hailo AI HAT+. Check out our Hackathon video How We Judged the Projects Our expert panel evaluated the projects based on several key criteria: Diversity & Creativity – Unique approaches and teamwork. Extra points were given to teams with members from various departments, such as marketing, design, and product management, showcasing the value of diverse perspectives and non-coding skills. Real-World Impact – Potential applications and benefits in everyday life. Implementation Quality – Code efficiency, performance, and robustness. Diversity & Creativity – Unique approaches and teamwork. Extra points were given to teams with members from various departments, such as marketing, design, and product management, showcasing the value of diverse perspectives and non-coding skills. Hackathon Takeaways The Hailo Hackathon 2024-2025 demonstrated that powerful AI applications can be developed in just 24 hours with the right tools. The Hailo AI HAT+ for Raspberry Pi provided teams... --- - Published: 2025-03-12 - Modified: 2025-06-17 - URL: https://hailo.ai/zh-hans/blog/hailo-hackathon-2024-2025-pushing-the-limits-of-ai-innovation-on-raspberry-pi/ - Categories: AI Hardware, Compute, Developer, Edge AI Device, Generative AI, Raspberry Pi 5 with Hailo AI HAT+ - Translation Priorities: 可选 The third annual Hailo Hackathon was bigger, bolder, and more innovative than ever! Over 24 hours, 60 Hailo employees came together to push the boundaries of edge AI using the Hailo AI HAT+ (26TOPS) on the Raspberry Pi 5. This wasn’t just a coding event—it was a celebration of creativity, collaboration, and problem-solving. With great food, an overnight coding marathon, and an electric atmosphere of team spirit, this hackathon proved that AI innovation can be both exciting and impactful. Participants developed real-world AI applications that are now available for the community to explore and expand upon. You can find the projects in our Community Projects. Note: Some of the projects are built to work only with the 26 TOPS Hailo AI HAT+. Check out our Hackathon video How We Judged the Projects Our expert panel evaluated the projects based on several key criteria: Diversity & Creativity – Unique approaches and teamwork. Extra points were given to teams with members from various departments, such as marketing, design, and product management, showcasing the value of diverse perspectives and non-coding skills. Real-World Impact – Potential applications and benefits in everyday life. Implementation Quality – Code efficiency, performance, and robustness. Diversity & Creativity – Unique approaches and teamwork. Extra points were given to teams with members from various departments, such as marketing, design, and product management, showcasing the value of diverse perspectives and non-coding skills. Hackathon Takeaways The Hailo Hackathon 2024-2025 demonstrated that powerful AI applications can be developed in just 24 hours with the right tools. The Hailo AI HAT+ for Raspberry Pi provided teams... --- > From zero-code robotics to advanced surveillance, see how Hailo’s Edge AI demos dazzled CES 2025. Explore the latest AI trends & highlights! - Published: 2025-02-04 - Modified: 2025-06-17 - URL: https://hailo.ai/blog/ces-2025-the-year-edge-ai-took-over/ - Categories: AI Hardware, Automotive, Compute, Developer, Edge AI Device, Generative AI, Intelligent Camera, Security, Surveillance, VMS - Translation Priorities: Optional Walking through the halls of CES 2025, it was impossible to ignore the dominance of AI in nearly every consumer product and technology. This year marked a true inflection point - AI is no longer confined to cloud computing or enterprise solutions; it has become an integral part of everyday life. From smart homes to personal assistants, from robots of all kinds to wearables like smartwatches and AR glasses, AI-powered devices are now more intelligent, responsive, and seamlessly integrated than ever before. One of the biggest highlights at CES this year was the rapid adoption of edge AI - the ability to process AI workloads directly on the device rather than relying on cloud computing. This trend isn’t just about making devices smarter; it’s about making them faster, more reliable, and more private. The advantages are clear: Always Available – Since processing happens locally, devices continue to function even without an internet connection. Real-Time Performance – Processing at the edge eliminates cloud-based latency, ensuring instant responsiveness. User Privacy – Sensitive data, including images and voice commands, remains on the device rather than being sent to the cloud. Hailo at CES 2025: Bringing Edge AI to Life With edge AI taking center stage, Hailo was at the heart of the action, demonstrating how its advanced AI processors enable cutting-edge applications across various industries, helping developers push the boundaries of what’s possible in AI-powered consumer electronics. At the Hailo booth, visitors explored a range of exciting live demos showcasing the versatility and... --- - Published: 2025-02-04 - Modified: 2025-06-17 - URL: https://hailo.ai/zh-hans/blog/ces-2025-the-year-edge-ai-took-over/ - Categories: AI Hardware, Automotive, Compute, Developer, Edge AI Device, Generative AI, Intelligent Camera, Security, Surveillance, VMS - Translation Priorities: 可选 Walking through the halls of CES 2025, it was impossible to ignore the dominance of AI in nearly every consumer product and technology. This year marked a true inflection point - AI is no longer confined to cloud computing or enterprise solutions; it has become an integral part of everyday life. From smart homes to personal assistants, from robots of all kinds to wearables like smartwatches and AR glasses, AI-powered devices are now more intelligent, responsive, and seamlessly integrated than ever before. One of the biggest highlights at CES this year was the rapid adoption of edge AI - the ability to process AI workloads directly on the device rather than relying on cloud computing. This trend isn’t just about making devices smarter; it’s about making them faster, more reliable, and more private. The advantages are clear: Always Available – Since processing happens locally, devices continue to function even without an internet connection. Real-Time Performance – Processing at the edge eliminates cloud-based latency, ensuring instant responsiveness. User Privacy – Sensitive data, including images and voice commands, remains on the device rather than being sent to the cloud. Hailo at CES 2025: Bringing Edge AI to Life With edge AI taking center stage, Hailo was at the heart of the action, demonstrating how its advanced AI processors enable cutting-edge applications across various industries, helping developers push the boundaries of what’s possible in AI-powered consumer electronics. At the Hailo booth, visitors explored a range of exciting live demos showcasing the versatility and... --- - Published: 2025-02-04 - Modified: 2025-06-17 - URL: https://hailo.ai/de/blog/ces-2025-the-year-edge-ai-took-over/ - Categories: AI Hardware, Automotive, Compute, Developer, Edge AI Device, Generative AI, Intelligent Camera, Security, Surveillance, VMS - Translation Priorities: Optional Walking through the halls of CES 2025, it was impossible to ignore the dominance of AI in nearly every consumer product and technology. This year marked a true inflection point - AI is no longer confined to cloud computing or enterprise solutions; it has become an integral part of everyday life. From smart homes to personal assistants, from robots of all kinds to wearables like smartwatches and AR glasses, AI-powered devices are now more intelligent, responsive, and seamlessly integrated than ever before. One of the biggest highlights at CES this year was the rapid adoption of edge AI - the ability to process AI workloads directly on the device rather than relying on cloud computing. This trend isn’t just about making devices smarter; it’s about making them faster, more reliable, and more private. The advantages are clear: Always Available – Since processing happens locally, devices continue to function even without an internet connection. Real-Time Performance – Processing at the edge eliminates cloud-based latency, ensuring instant responsiveness. User Privacy – Sensitive data, including images and voice commands, remains on the device rather than being sent to the cloud. Hailo at CES 2025: Bringing Edge AI to Life With edge AI taking center stage, Hailo was at the heart of the action, demonstrating how its advanced AI processors enable cutting-edge applications across various industries, helping developers push the boundaries of what’s possible in AI-powered consumer electronics. At the Hailo booth, visitors explored a range of exciting live demos showcasing the versatility and... --- - Published: 2025-02-04 - Modified: 2025-06-17 - URL: https://hailo.ai/ja/blog/ces-2025-the-year-edge-ai-took-over/ - Categories: AI Hardware, Automotive, Compute, Developer, Edge AI Device, Generative AI, Intelligent Camera, Security, Surveillance, VMS - Translation Priorities: 可选 Walking through the halls of CES 2025, it was impossible to ignore the dominance of AI in nearly every consumer product and technology. This year marked a true inflection point - AI is no longer confined to cloud computing or enterprise solutions; it has become an integral part of everyday life. From smart homes to personal assistants, from robots of all kinds to wearables like smartwatches and AR glasses, AI-powered devices are now more intelligent, responsive, and seamlessly integrated than ever before. One of the biggest highlights at CES this year was the rapid adoption of edge AI - the ability to process AI workloads directly on the device rather than relying on cloud computing. This trend isn’t just about making devices smarter; it’s about making them faster, more reliable, and more private. The advantages are clear: Always Available – Since processing happens locally, devices continue to function even without an internet connection. Real-Time Performance – Processing at the edge eliminates cloud-based latency, ensuring instant responsiveness. User Privacy – Sensitive data, including images and voice commands, remains on the device rather than being sent to the cloud. Hailo at CES 2025: Bringing Edge AI to Life With edge AI taking center stage, Hailo was at the heart of the action, demonstrating how its advanced AI processors enable cutting-edge applications across various industries, helping developers push the boundaries of what’s possible in AI-powered consumer electronics. At the Hailo booth, visitors explored a range of exciting live demos showcasing the versatility and... --- > Elevate your security with Hailo's scalable AI-driven video management systems for unparalleled surveillance performance. - Published: 2025-01-31 - Modified: 2025-04-17 - URL: https://hailo.ai/blog/ai-enhanced-video-management-scalability/ - Categories: Security, VMS - Tags: hailo-8-ai-accelerator, hailo-8-century, VMS - Translation Priorities: Optional https://open. spotify. com/episode/5H3fryaiMusbi2wLVwVvir? si=9bfa1edda3104c76 What is a video management system? Video Management Systems, also known as VMS, collect inputs from multiple cameras and other sensors, addressing all related aspects of video handling, such as storage, retrieval, analysis and display. VMS are typically used in the security and surveillance space, enhancing personal safety in public areas, office buildings, transportation terminals, medical institutes, and more. Other typical uses include the extraction of business intelligence through user behavior analysis for the purpose of customer experience improvement in retail and other industries.  Traditionally, the analysis of multiple video streams used to be laborious, relying on human perception for visual identification of events happening across a multitude of video feeds. This method has many disadvantages as it is difficult to scale, violates people’s privacy, and is prone to errors due to operator’s fatigue, leading to false alarms, missed occurrences, and inefficient use of resources.  Nowadays, deep learning is enabling the automation of video analytics tasks, thereby allowing for scalability, and improvement in overall performance. This eventually leads to lower total cost of ownership (TCO).   According to a recent market research, the video management system market size is expected to reach $31B by 2027, growing at a Compound Annual Growth Rate (CAGR) of 23. 1% between 2022-2027. The key drivers for this growth are increasing security concerns and rapid adoption of IP cameras for surveillance, security, and retail applications.   AI-powered video analytics are being rapidly adopted for VMS  There are multiple possible configurations to a... --- - Published: 2025-01-31 - Modified: 2025-01-09 - URL: https://hailo.ai/de/blog/ai-enhanced-video-management-scalability/ - Categories: Security, VMS - Tags: hailo-8-ai-accelerator, hailo-8-century, VMS - Translation Priorities: Optional What is a video management system? Video Management Systems, also known as VMS, collect inputs from multiple cameras and other sensors, addressing all related aspects of video handling, such as storage, retrieval, analysis and display. VMS are typically used in the security and surveillance space, enhancing personal safety in public areas, office buildings, transportation terminals, medical institutes, and more. Other typical uses include the extraction of business intelligence through user behavior analysis for the purpose of customer experience improvement in retail and other industries.  Traditionally, the analysis of multiple video streams used to be laborious, relying on human perception for visual identification of events happening across a multitude of video feeds. This method has many disadvantages as it is difficult to scale, violates people’s privacy, and is prone to errors due to operator’s fatigue, leading to false alarms, missed occurrences, and inefficient use of resources.  Nowadays, deep learning is enabling the automation of video analytics tasks, thereby allowing for scalability, and improvement in overall performance. This eventually leads to lower total cost of ownership (TCO).   According to a recent market research, the video management system market size is expected to reach $31B by 2027, growing at a Compound Annual Growth Rate (CAGR) of 23. 1% between 2022-2027. The key drivers for this growth are increasing security concerns and rapid adoption of IP cameras for surveillance, security, and retail applications.   AI-powered video analytics are being rapidly adopted for VMS  There are multiple possible configurations to a VMS system, depending on... --- > Hailo のスケーラブルな AI 駆動ビデオ管理システムでセキュリティを強化し、比類のない監視パフォーマンスを実現します。 - Published: 2025-01-31 - Modified: 2024-08-25 - URL: https://hailo.ai/ja/blog/ai-enhanced-video-management-scalability/ - Categories: Security, VMS - Tags: hailo-8-ai-accelerator, hailo-8-century, VMS - Translation Priorities: 可选 Video Management Systems, also known as VMS, collect inputs from multiple cameras and other sensors, addressing all related aspects of video handling, such as storage, retrieval, analysis and display. VMS are typically used in the security and surveillance space, enhancing personal safety in public areas, office buildings, transportation terminals, medical institutes, and more. Other typical uses include the extraction of business intelligence through user behavior analysis for the purpose of customer experience improvement in retail and other industries. Traditionally, the analysis of multiple video streams used to be laborious, relying on human perception for visual identification of events happening across a multitude of video feeds. This method has many disadvantages as it is difficult to scale, violates people’s privacy, and is prone to errors due to operator’s fatigue, leading to false alarms, missed occurrences, and inefficient use of resources. Nowadays, deep learning is enabling the automation of video analytics tasks, thereby allowing for scalability, and improvement in overall performance. This eventually leads to lower total cost of ownership (TCO). According to a recent market research, the video management system market size is expected to reach $31B by 2027, growing at a Compound Annual Growth Rate (CAGR) of 23. 1% between 2022-2027. The key drivers for this growth are increasing security concerns and rapid adoption of IP cameras for surveillance, security, and retail applications. AI-powered video analytics are being rapidly adopted for VMS There are multiple possible configurations to a VMS system, depending on the number of video channels, the required... --- - Published: 2025-01-31 - Modified: 2025-01-09 - URL: https://hailo.ai/zh-hans/blog/ai-enhanced-video-management-scalability/ - Categories: Security, VMS - Tags: hailo-8-ai-accelerator, hailo-8-century, VMS - Translation Priorities: 可选 What is a video management system? Video Management Systems, also known as VMS, collect inputs from multiple cameras and other sensors, addressing all related aspects of video handling, such as storage, retrieval, analysis and display. VMS are typically used in the security and surveillance space, enhancing personal safety in public areas, office buildings, transportation terminals, medical institutes, and more. Other typical uses include the extraction of business intelligence through user behavior analysis for the purpose of customer experience improvement in retail and other industries.  Traditionally, the analysis of multiple video streams used to be laborious, relying on human perception for visual identification of events happening across a multitude of video feeds. This method has many disadvantages as it is difficult to scale, violates people’s privacy, and is prone to errors due to operator’s fatigue, leading to false alarms, missed occurrences, and inefficient use of resources.  Nowadays, deep learning is enabling the automation of video analytics tasks, thereby allowing for scalability, and improvement in overall performance. This eventually leads to lower total cost of ownership (TCO).   According to a recent market research, the video management system market size is expected to reach $31B by 2027, growing at a Compound Annual Growth Rate (CAGR) of 23. 1% between 2022-2027. The key drivers for this growth are increasing security concerns and rapid adoption of IP cameras for surveillance, security, and retail applications.   AI-powered video analytics are being rapidly adopted for VMS  There are multiple possible configurations to a VMS system, depending on... --- > Explore how AI in public safety balances security & privacy. Our edge AI solutions ensure secure environments while respecting privacy. Learn how! - Published: 2024-12-20 - Modified: 2025-02-03 - URL: https://hailo.ai/blog/ai-in-public-safety-privacy-and-security/ - Categories: Edge AI Device, Intelligent Camera, Privacy, Security, Smart City, Surveillance - Translation Priorities: Optional https://open. spotify. com/episode/4q5Bh7p9SwVWnQY3RYww96? si=53dcd6b4f0394151 The Clash of Safety and Privacy  In today's world, rising urbanization, increasing crime rates, and the threat of terrorism are putting public safety at risk. As cities expand and their population grows denser, the challenge of ensuring public safety becomes even more complex, especially considering constrained law enforcement resources. Advances in technology have led to the deployment of monitoring devices and cameras, to make public spaces safer. With an installed base of over 600 million surveillance cameras, China has almost one camera per 2 people. The top most surveilled cities outside of China include Delhi, Seoul, Moscow, New York and London among others. However, this increase in surveillance comes at a significant cost: the erosion of personal privacy. People value their right to remain anonymous and free from constant monitoring. The feeling that the "Big Brother" is watching at all times is leading to a complex clash between safety and privacy. This triggers a vivid debate among policy makers, that often leads to legislation to regulate or prohibit the use of monitoring devices in the public domain.   Figure 1 - Reference: https://www. comparitech. com/vpn-privacy/the-worlds-most-surveilled-cities/ AI Technologies for Crime Prevention and Enhanced Public Safety  AI has been playing a growing role in maintaining public safety recently, through integration into security systems at the camera or the video management system level. Advancement in technology, especially around generative AI, makes AI even more attractive for public safety monitoring. The most common AI use cases in surveillance systems include perimeter... --- > 探索公共安全领域的人工智能如何平衡安全性和隐私性。我们的边缘人工智能解决方案在确保安全环境的同时尊重隐私。了解如何做到！ - Published: 2024-12-20 - Modified: 2024-08-25 - URL: https://hailo.ai/zh-hans/blog/ai-in-public-safety-privacy-and-security/ - Categories: Edge AI Device, Intelligent Camera, Privacy, Security, Smart City, Surveillance - Translation Priorities: 可选 安全与隐私的冲突  当今世界，不断加快的城市化进程、不断上升的犯罪率和恐怖主义威胁使得公共安全面临着风险。随着城市的扩张和人口的密集，确保公共安全的挑战变得更加复杂，在执法资源有限的情况下尤其如此。技术的进步推动了监控设备和摄像头的部署，使公共场所变得更加安全。中国安装了6亿多个监控摄像头，几乎每两个人就有一个摄像头。在中国以外地区，监控设备最多的城市包括德里、首尔、莫斯科、纽约和伦敦等。然而，监控的增加也付出了巨大的代价：个人隐私受到侵蚀。人们重视自己匿名和不受持续监控的权利。“老大哥”时时刻刻都在监视的感觉正导致安全与隐私之间出现复杂的冲突。这在政策制定者中引发了一场生动的辩论，往往会导致通过立法来规范或禁止在公共场所使用监控设备。  图1 - 参考资料: https://www. comparitech. com/vpn-privacy/the-worlds-most-surveilled-cities/ 人工智能技术用于预防犯罪和加强公共安全 通过在摄像头或视频管理系统层面集成安全系统，人工智能在维护公共安全方面发挥着越来越重要的作用。技术进步，尤其是生成式人工智能领域的进步，使人工智能对公共安全监控更具吸引力。 监控系统中最常见的人工智能用例包括周边保护和访问控制。这些应用利用了人工智能任务，例如物体检测、分割、视频元数据和重识别，快速准确地识别合法、可疑或异常的人员或行为，并触发实时响应。 人工智能驱动的监控系统可提供更精密、更细微的监控功能，支持对安全事件做出实时且高精度的检测、识别和响应。然而，在增强安全和确保公共安全的同时，这些技术也引起了人们对隐私和个人身份信息（PII）可能被滥用的担忧，突出了采取强有力的数据保护措施的必要性。云人工智能解决方案的隐私挑战。 云人工智能解决方案的伦理考虑因素和隐私挑战 传统的云端人工智能解决方案利用集中式数据中心提供强大的处理能力。不过，它们也带来了一些漏洞，特别是在数据隐私方面： 静态数据:集中存储大量数据使得云系统成为网络攻击的目标。黑客，无论是个人、有组织犯罪集团，甚至是敌对政府，都可能利用这些系统，导致大规模数据泄露。将数据处理分散到网络边缘，使得任何漏洞都仅限于被黑客攻击的特定节点，从而有效避免大规模数据泄露。此外，有关数据隐私的法规也限制了分析原始数据的方式和范围。云端系统必须应对这些复杂的法律环境，这往往会限制洞察力，带来合规性挑战，甚至导致潜在的法律责任。另一方面，边缘处理只需存储和传输所需的最低限度信息即可获得深刻的见解。 传输中的数据: 将数据从设备传输到云端会产生多个漏洞点。在传输过程中拦截数据会暴露敏感信息，破坏系统的安全性。 可信执行环境: 云中心是一个单点故障点，可能会对大量摄像头造成影响，而如果是分布式系统，每个系统都可自由采用不同的算法/功能，根据所有者/集成商的决策来提高精确度。 边缘人工智能：隐私敏感型安全解决方案 边缘人工智能在设备本地处理数据，而不是将数据传输到集中式云端，为应对这些挑战提供了令人信服的解决方案。从隐私角度看，这种方法有几个优点： 减少数据传输: 通过在设备上处理数据，边缘人工智能最大限度地减少了通过互联网传输敏感信息的需求，从而大大降低了数据被截获和泄露的风险。 本地数据存储: 边缘设备在本地存储数据，这就限制了网络攻击时的风险范围。即使一台设备被入侵，入侵范围也仅限于该特定设备，而不是整个网络. 匿名数据存储: 此外，如果在本地进行匿名处理，则可以对存储在边缘设备或云中的数据进行匿名处理，从而在不暴露个人身份信息的情况下保持数据的本质. 数据选择性: 边缘人工智能的设计可以只关注相关事件，如识别暴力事件或可疑行为，而无需记录连续镜头。这种选择性的记录方式有助于维护个人在公共场所的隐私. 为了有效平衡安全与隐私，边缘人工智能系统在设计上可以进行特定的限制，从而从根本上保护个人数据。例如，带宽限制会限制摄像头的传输能力，以确保视频文件不会持续发送到云端。这降低了数据泄露的风险，保护了个人隐私。另一种原生技术限制是选择性地记录，以限制存储数据的数量，只捕捉公共安全所需的数据. 边缘人工智能要想发挥效用，就必须既强大又高效。设备需要快速处理复杂的算法，以实时识别威胁，同时保持成本效益和能效。虽然独立软件开发商正在优化算法，以确保边缘人工智能能够在不耗费计算资源的情况下执行复杂的任务，但人工智能硬件（如专用人工智能处理器和低功耗的高性能芯片）的进步也使边缘人工智能成为可能. 寻求平衡：隐私和安全的统一 边缘人工智能为平衡公共安全与个人隐私的挑战提供了一个前景广阔的解决方案。边缘人工智能在本地处理数据并对数据传输和存储施加固有限制，降低了与云端系统相关的风险。随着这些技术的不断发展，边缘人工智能将在创建更安全的公共空间方面发挥关键作用，同时尊重个人的匿名权利。这种方法不仅增强了安全性，而且还建立了对那些旨在保护我们的系统的信任. --- > Entdecken Sie, wie KI in öffentlichen Sicherheitslösungen sichere Umgebungen unter Wahrung der Privatsphäre bietet. Erfahren Sie wie! - Published: 2024-12-20 - Modified: 2024-08-25 - URL: https://hailo.ai/de/blog/ai-in-public-safety-privacy-and-security/ - Categories: Edge AI Device, Intelligent Camera, Privacy, Security, Smart City, Surveillance - Translation Priorities: Optional Der Konflikt zwischen Sicherheit und Privatsphäre  Die zunehmende Verstädterung, die steigende Kriminalitätsrate und die Bedrohung durch den Terrorismus gefährden die öffentliche Sicherheit in der heutigen Welt. Mit der Ausdehnung der Städte und der zunehmenden Bevölkerungsdichte wird die Gewährleistung der öffentlichen Sicherheit zu einer noch komplexeren Herausforderung, insbesondere angesichts der begrenzten Ressourcen der Strafverfolgungsbehörden. Die Fortschritte in der Technologie haben zum Einsatz von Überwachungsgeräten und Kameras geführt, um öffentliche Räume sicherer zu machen. Mit einer installierten Basis von über 600 Millionen Überwachungskameras kommt in China fast eine Kamera auf 2 Einwohner. Zu den am meisten überwachten Städten außerhalb Chinas gehören unter anderem Delhi, Seoul, Moskau, New York und London. Diese zunehmende Überwachung hat jedoch einen hohen Preis: die Aushöhlung der persönlichen Privatsphäre. Die Menschen schätzen ihr Recht, anonym und frei von ständiger Überwachung zu bleiben. Das Gefühl, dass der "Große Bruder" uns ständig beobachtet, führt zu einem komplexen Konflikt zwischen Sicherheit und Privatsphäre. Dies löst eine lebhafte Debatte unter den politischen Entscheidungsträgern aus, die oft zu Gesetzen führt, die den Einsatz von Überwachungsgeräten im öffentlichen Bereich regeln oder verbieten.   Abbildung 1 - Referenz: https://www. comparitech. com/vpn-privacy/the-worlds-most-surveilled-cities/ KI-Technologien für die Verbrechensprävention und die Verbesserung der öffentlichen Sicherheit KI spielt in letzter Zeit eine immer wichtigere Rolle bei der Aufrechterhaltung der öffentlichen Sicherheit, indem sie in Sicherheitssysteme auf der Ebene der Kameras oder desintegriert wird. Der technologische Fortschritt, insbesondere im Bereich der generative KI macht KI für die Überwachung der öffentlichen Sicherheit noch attraktiver. Zu den häufigsten Anwendungsfällen von KI in Überwachungssystemen... --- > 公共の安全における AI がセキュリティとプライバシーのバランスをどのように保っているかをご覧ください。当社のエッジ AI ソリューションは、プライバシーを尊重しながら安全な環境を確保します。その方法をご覧ください。 - Published: 2024-12-20 - Modified: 2024-09-22 - URL: https://hailo.ai/ja/blog/ai-in-public-safety-privacy-and-security/ - Categories: Edge AI Device, Intelligent Camera, Privacy, Security, Smart City, Surveillance - Translation Priorities: 可选 安全とプライバシーの対立 現代社会では、都市化の進行、犯罪率の増加、そしてテロの脅威が公共の安全を脅かしています。都市が拡大し、人口密度が高まるにつれて、限られた法執行機関のリソースを考慮すると、公共の安全を確保するための課題はますます複雑化してきています。技術の進歩により、公共の場をより安全にするために、監視装置やカメラの導入が進んでいます。中国では、6億台以上の監視カメラが設置されており、ほぼ2人に1台のカメラが設置されていることになります。中国以外でも監視カメラが多い都市には、デリー、ソウル、モスクワ、ニューヨーク、ロンドンなどがあります。しかし、この監視の増加は、個人のプライバシーの侵害という大きな代償を伴います。人々は匿名性を維持し、絶え間ない監視から自由である権利を大切にしています。「ビッグブラザー」に常に見張られているという感覚は、安全とプライバシーの間で複雑な衝突を引き起こしています。このことは、政策立案者の間で激しい議論を引き起こし、しばしば公共の場での監視装置の使用を規制または禁止する立法に繋がっています。 図1 - 参照: https://www. comparitech. com/vpn-privacy/the-worlds-most-surveilled-cities/ AI技術による犯罪防止と公共安全の向上 AIは近年、カメラやビデオ管理システムのレベルでセキュリティシステムに統合され、公共の安全を維持するための役割がますます重要になっています。特に生成AIの技術が進化することで、AIは公共安全の監視にとってさらに魅力的なものとなっています。 監視システムにおける最も一般的なAIの使用例には、周辺保護とアクセス制御が含まれます。これらのアプリケーションは、物体検出、セグメンテーション、ビデオメタデータ、および再識別などのAIタスクを利用して、正当な人物や行動と、不審で異常な人物や行動とを、迅速かつ正確に識別し、リアルタイムで対応を開始します。 AI搭載の監視システムは、より洗練された繊細な監視能力を提供し、リアルタイムでのセキュリティ上の問題の検出、識別、および対応を可能にします。ただし、セキュリティを強化し公共の安全を確保する一方で、これらの技術は個人情報のプライバシーや個人を特定できる情報の悪用に関する懸念を引き起こし、強固なデータ保護措置の必要性を浮き彫りにします。これが、クラウドAIソリューションにおけるプライバシーの課題です。 クラウドAIソリューションにおける倫理的考慮とプライバシーの課題 従来のクラウドベースのAIソリューションは、中央集中型データセンターを活用して強力な処理能力を提供しますが、同時にデータプライバシーに関するいくつかの脆弱性ももたらします 保存データ：大量のデータを集中管理することは、クラウドシステムをサイバー攻撃の格好のターゲットになりえます。ハッカー、組織犯罪組織、さらには敵対的な政府などがこれらのシステムを悪用し、大規模なデータ漏洩を引き起こす可能性があります。データ処理をネットワークのエッジに分散させることで、特定のノードがハッキングされた場合の被害を最小化し、大規模なデータ漏洩を困難にします。さらに、データプライバシーに関する規制は、生データの分析方法や内容に制約を課すため、クラウドベースのシステムは複雑な法的状況を乗り越える必要があり、これが洞察の制限やコンプライアンスの課題、さらには法的責任の可能性を引き起こすことがあります。一方、エッジ処理は、必要最小限の情報だけを保存し送信することで、深い洞察を収集することが可能です。 転送中のデータ: デバイスからクラウドへのデータ転送は、複数の脆弱性ポイントを生み出します。転送中にデータが傍受されると、システムのセキュリティを損なう可能性があり、機密情報が漏洩する危険があります。 信頼できる実行環境: クラウドセンターは、多数のカメラに影響を与える可能性のある単一障害点となりえます。一方で、分散型のシステムでは、各システムが所有者やインテグレーターの判断に基づいて、精度を向上させるさまざまなアルゴリズムや機能を採用できるという、自由があります。 エッジAI：プライバシー重視のセキュリティのためのソリューション Edge AIは、データを中央のクラウドに送信するのではなく、デバイス自体でローカルに処理することで、これらの課題に対する魅力的な解決策を提供します。このアプローチは、プライバシーの観点からいくつかの利点をもたらします データ通信量の削減:エッジAIはデバイス上でデータを処理することで、インターネット上での機密情報の転送を最小限に抑え、傍受や侵害のリスクを大幅に減少させます。 ローカライズされたデータ保存:エッジデバイスはデータをローカルに保存するため、サイバー攻撃が発生した場合でも被害の範囲が限定されます。仮にデバイスが侵害されても、被害はその特定のデバイスに限定され、ネットワーク全体に及びません。 匿名化されたデータ保存:さらに、匿名化がローカルで行われる場合、エッジデバイスやクラウドに保存されたデータは匿名化され、個人情報を公開することなくデータの本質を保持することが可能です。 データ選別:エッジAIは、暴力や不審な行動の特定など、関連するイベントのみに焦点を当てるように設計することができ、連続的な映像の記録を避けることができます。この選択的な記録により、公共の場での個人のプライバシーを保護することができます。 安全性とプライバシーのバランスを効果的に取るために、エッジAIシステムは、個人データを保護するための特定の制限を設けるよう設計できます。例えば、カメラの転送機能を制限する帯域幅制限を設定することで、ビデオファイルが継続的にクラウドに送信されるのを防ぎます。これにより、データ侵害のリスクを軽減し、個人のプライバシーが保護されます。また、もうひとつのネイティブ技術の制限として、選択的な記録を適用し、保存されるデータ量を減らし、公共の安全に必要なものだけを記録することも可能です。 エッジAIが効果的であるためには、強力で効率的である必要があります。デバイスは、リアルタイムで脅威を特定するために複雑なアルゴリズムを迅速に処理する一方で、コスト効率や電力効率を維持する必要があります。独立系ソフトウェアベンダー（ISV）は、エッジAIが計算リソースを消費することなく高度なタスクを実行できるようにアルゴリズムを最適化しており、AIハードウェアの進歩、例えば、特殊なAIプロセッサや低電力・高性能チップがエッジAIの実現を可能にしています。 バランスを取る：プライバシーと安全性の両立 エッジAIは、公共の安全と個人のプライバシーのバランスを取るという課題に対する有望な解決策を提供します。データをローカルで処理し、データの転送や保存に制限を課すことで、エッジAIはクラウドベースのシステムに関連するリスクを軽減します。これらの技術が進化し続ける中、エッジAIは個人の匿名性を尊重しながら公共の場をより安全にする上で重要な役割を果たすでしょう。このアプローチは、セキュリティを強化するだけでなく、私たちを守るために設計されたシステムへの信頼も築きます。 --- > Innovating at the technological forefront with Generative AI at the Edge. Learn about its role in the evolution of modern edge computing. - Published: 2024-12-08 - Modified: 2025-03-24 - URL: https://hailo.ai/blog/the-evolution-of-ai-on-the-edge-from-perception-to-creation/ - Categories: AI Hardware, Compute, Edge AI Device, Generative AI - Translation Priorities: Optional In the vast landscape of artificial intelligence (AI), one of the most intriguing journeys has been the evolution of AI on the edge. This journey has taken us from classic machine vision to the realms of discriminative AI, enhancive AI, and now, the groundbreaking frontier of generative AI. Each step has brought us closer to a future where intelligent systems seamlessly integrate with our daily lives, offering an immersive experience of not just perception but also creation at the palm of our hand. In the vast landscape of artificial intelligence (AI), one of the most intriguing journeys has been the evolution of AI on the edge. This journey has taken us from classic machine vision to the realms of perceptive AI, enhancive AI, and now, the groundbreaking frontier of generative AI. Each step has brought us closer to a future where intelligent systems seamlessly integrate with our daily lives, offering an immersive experience of not just perception but also creation at the palm of our hand. From Perception to Understanding: The Rise of Perceptive AI The journey began with machine vision, enabling computers to perceive and interpret the visual world around them. However, it was the advent of AI-powered video analytics that truly revolutionized this field. Perceptive AI empowered machines to not only recognize objects and scenes but also understand them. Tasks like object detection, and instance segmentation have long surpassed human performance level, thereby enabling machines to identify individuals, vehicles, animals, and other objects reliably and trigger real-time activity accordingly. This is successfully put into use in surveillance, safety monitoring, and law enforcement applications. Enhancing Perception: The Emergence of Enhancive AI With the growing understanding of the theory of neural network operation, and the successful results they have yielded, its application widened beyond just information retrieval and extraction. Using semantic understanding of the physical nature of the scene, neural networks can be leveraged to enhance the visual quality of images, for more pleasing results as well as to further enhance perception and... --- - Published: 2024-12-08 - Modified: 2024-08-25 - URL: https://hailo.ai/zh-hans/blog/the-evolution-of-ai-on-the-edge-from-perception-to-creation/ - Categories: AI Hardware, Compute, Edge AI Device, Generative AI, Generative AI - Translation Priorities: 可选 In the vast landscape of artificial intelligence (AI), one of the most intriguing journeys has been the evolution of AI on the edge. This journey has taken us from classic machine vision to the realms of discriminative AI, enhancive AI, and now, the groundbreaking frontier of generative AI. Each step has brought us closer to a future where intelligent systems seamlessly integrate with our daily lives, offering an immersive experience of not just perception but also creation at the palm of our hand. In the vast landscape of artificial intelligence (AI), one of the most intriguing journeys has been the evolution of AI on the edge. This journey has taken us from classic machine vision to the realms of discriminative AI, enhancive AI, and now, the groundbreaking frontier of generative AI. Each step has brought us closer to a future where intelligent systems seamlessly integrate with our daily lives, offering an immersive experience of not just perception but also creation at the palm of our hand. From Perception to Understanding: The Rise of Discriminative AI The journey began with machine vision, enabling computers to perceive and interpret the visual world around them. However, it was the advent of AI-powered video analytics that truly revolutionized this field. Discriminative AI empowered machines to not only recognize objects and scenes but also understand them. Tasks like object detection, and instance segmentation have long surpassed human performance level, thereby enabling machines to identify individuals, vehicles, animals, and other objects reliably and trigger real-time activity accordingly. This is successfully put into use in surveillance, safety monitoring, and law enforcement applications. Enhancing Perception: The Emergence of Enhancive AI With the growing understanding of the theory of neural network operation, and the successful results they have yielded, its application widened beyond just information retrieval and extraction. Using semantic understanding of the physical nature of the scene, neural networks can be leveraged to enhance the visual quality of images, for more pleasing results as well as to further enhance perception and... --- - Published: 2024-12-08 - Modified: 2025-03-24 - URL: https://hailo.ai/de/blog/the-evolution-of-ai-on-the-edge-from-perception-to-creation/ - Categories: AI Hardware, Compute, Edge AI Device, Generative AI - Translation Priorities: Optional In the vast landscape of artificial intelligence (AI), one of the most intriguing journeys has been the evolution of AI on the edge. This journey has taken us from classic machine vision to the realms of discriminative AI, enhancive AI, and now, the groundbreaking frontier of generative AI. Each step has brought us closer to a future where intelligent systems seamlessly integrate with our daily lives, offering an immersive experience of not just perception but also creation at the palm of our hand. In the vast landscape of artificial intelligence (AI), one of the most intriguing journeys has been the evolution of AI on the edge. This journey has taken us from classic machine vision to the realms of perceptive AI, enhancive AI, and now, the groundbreaking frontier of generative AI. Each step has brought us closer to a future where intelligent systems seamlessly integrate with our daily lives, offering an immersive experience of not just perception but also creation at the palm of our hand. From Perception to Understanding: The Rise of Perceptive AI The journey began with machine vision, enabling computers to perceive and interpret the visual world around them. However, it was the advent of AI-powered video analytics that truly revolutionized this field. Perceptive AI empowered machines to not only recognize objects and scenes but also understand them. Tasks like object detection, and instance segmentation have long surpassed human performance level, thereby enabling machines to identify individuals, vehicles, animals, and other objects reliably and trigger real-time activity accordingly. This is successfully put into use in surveillance, safety monitoring, and law enforcement applications. Enhancing Perception: The Emergence of Enhancive AI With the growing understanding of the theory of neural network operation, and the successful results they have yielded, its application widened beyond just information retrieval and extraction. Using semantic understanding of the physical nature of the scene, neural networks can be leveraged to enhance the visual quality of images, for more pleasing results as well as to further enhance perception and... --- - Published: 2024-12-08 - Modified: 2025-03-24 - URL: https://hailo.ai/ja/blog/the-evolution-of-ai-on-the-edge-from-perception-to-creation/ - Categories: AI Hardware, Compute, Edge AI Device, Generative AI - Translation Priorities: 可选 In the vast landscape of artificial intelligence (AI), one of the most intriguing journeys has been the evolution of AI on the edge. This journey has taken us from classic machine vision to the realms of discriminative AI, enhancive AI, and now, the groundbreaking frontier of generative AI. Each step has brought us closer to a future where intelligent systems seamlessly integrate with our daily lives, offering an immersive experience of not just perception but also creation at the palm of our hand. In the vast landscape of artificial intelligence (AI), one of the most intriguing journeys has been the evolution of AI on the edge. This journey has taken us from classic machine vision to the realms of perceptive AI, enhancive AI, and now, the groundbreaking frontier of generative AI. Each step has brought us closer to a future where intelligent systems seamlessly integrate with our daily lives, offering an immersive experience of not just perception but also creation at the palm of our hand. From Perception to Understanding: The Rise of Perceptive AI The journey began with machine vision, enabling computers to perceive and interpret the visual world around them. However, it was the advent of AI-powered video analytics that truly revolutionized this field. Perceptive AI empowered machines to not only recognize objects and scenes but also understand them. Tasks like object detection, and instance segmentation have long surpassed human performance level, thereby enabling machines to identify individuals, vehicles, animals, and other objects reliably and trigger real-time activity accordingly. This is successfully put into use in surveillance, safety monitoring, and law enforcement applications. Enhancing Perception: The Emergence of Enhancive AI With the growing understanding of the theory of neural network operation, and the successful results they have yielded, its application widened beyond just information retrieval and extraction. Using semantic understanding of the physical nature of the scene, neural networks can be leveraged to enhance the visual quality of images, for more pleasing results as well as to further enhance perception and... --- > The future of smart cameras with Hailo technologies combines AI algorithms with real-time smart video analytics for various applications. - Published: 2024-11-09 - Modified: 2025-03-24 - URL: https://hailo.ai/blog/ai-cameras-from-vision-to-insights/ - Categories: Edge AI Developer, Edge AI Device, Intelligent Camera - Translation Priorities: Optional As the camera market is booming, so does the need to empower cameras with artificial intelligence (AI). In this blog post we discuss the need for on-camera AI and how it enhances video quality and empowers advanced video analytics. We will look at some typical cameras and applications and try to estimate the AI budget required to execute the different scenarios. https://open. spotify. com/episode/5Y3OvYHSu8QjCYn1Dl9pZ9? si=d9a7dbfc9f9c475a The Essential Role of AI-Powered Cameras in Shaping a Smarter WorldAs the camera market is booming, so does the need to empower cameras with artificial intelligence (AI). In this blog post we discuss the need for on-camera AI and how it enhances video quality and empowers advanced video analytics. We will look at some typical cameras and applications and try to estimate the AI budget required to execute the different scenarios. Cameras. Cameras Everywhere In today’s tech-driven world, cameras have become an integral part of our daily lives, and we are used to constantly recording videos and being recorded. The rapid deployment of IP cameras in residential homes, commercial and public space and the industrial sector, is fueling an unprecedented growth in the market, which is estimated by ABI Research to reach 200 million cameras by 2027, with a revenue of $35B. The most significant growth driver for this market is the ability to improve safety and security through video surveillance. From enhancing home security to monitoring public safety and optimizing traffic management, smart cameras provide countless benefits and endless opportunities for a smarter, safer world. AI isn’t only cloud anymore With the proliferation of camera deployments, comes the need to automate and enhance the ability to monitor the video streams and generate insights from them, as well as to make streaming and storage of video more efficient and cost-effective. This is where artificial intelligence (AI) comes to play. Since traditional cloud-based AI models often suffer... --- - Published: 2024-11-09 - Modified: 2025-02-17 - URL: https://hailo.ai/de/blog/ai-cameras-from-vision-to-insights/ - Categories: Edge AI Developer, Edge AI Device, Intelligent Camera - Translation Priorities: Optional As the camera market is booming, so does the need to empower cameras with artificial intelligence (AI). In this blog post we discuss the need for on-camera AI and how it enhances video quality and empowers advanced video analytics. We will look at some typical cameras and applications and try to estimate the AI budget required to execute the different scenarios. The Essential Role of AI-Powered Cameras in Shaping a Smarter WorldAs the camera market is booming, so does the need to empower cameras with artificial intelligence (AI). In this blog post we discuss the need for on-camera AI and how it enhances video quality and empowers advanced video analytics. We will look at some typical cameras and applications and try to estimate the AI budget required to execute the different scenarios. Cameras. Cameras Everywhere In today’s tech-driven world, cameras have become an integral part of our daily lives, and we are used to constantly recording videos and being recorded. The rapid deployment of IP cameras in residential homes, commercial and public space and the industrial sector, is fueling an unprecedented growth in the market, which is estimated by ABI Research to reach 200 million cameras by 2027, with a revenue of $35B. The most significant growth driver for this market is the ability to improve safety and security through video surveillance. From enhancing home security to monitoring public safety and optimizing traffic management, smart cameras provide countless benefits and endless opportunities for a smarter, safer world. AI isn’t only cloud anymore With the proliferation of camera deployments, comes the need to automate and enhance the ability to monitor the video streams and generate insights from them, as well as to make streaming and storage of video more efficient and cost-effective. This is where artificial intelligence (AI) comes to play. Since traditional cloud-based AI models often suffer from latency issues, not... --- - Published: 2024-11-09 - Modified: 2025-01-09 - URL: https://hailo.ai/zh-hans/blog/ai-cameras-from-vision-to-insights/ - Categories: Edge AI Developer, Edge AI Device, Intelligent Camera - Translation Priorities: 可选 As the camera market is booming, so does the need to empower cameras with artificial intelligence (AI). In this blog post we discuss the need for on-camera AI and how it enhances video quality and empowers advanced video analytics. We will look at some typical cameras and applications and try to estimate the AI budget required to execute the different scenarios. The Essential Role of AI-Powered Cameras in Shaping a Smarter WorldAs the camera market is booming, so does the need to empower cameras with artificial intelligence (AI). In this blog post we discuss the need for on-camera AI and how it enhances video quality and empowers advanced video analytics. We will look at some typical cameras and applications and try to estimate the AI budget required to execute the different scenarios. Cameras. Cameras Everywhere In today’s tech-driven world, cameras have become an integral part of our daily lives, and we are used to constantly recording videos and being recorded. The rapid deployment of IP cameras in residential homes, commercial and public space and the industrial sector, is fueling an unprecedented growth in the market, which is estimated by ABI Research to reach 200 million cameras by 2027, with a revenue of $35B. The most significant growth driver for this market is the ability to improve safety and security through video surveillance. From enhancing home security to monitoring public safety and optimizing traffic management, smart cameras provide countless benefits and endless opportunities for a smarter, safer world. AI isn’t only cloud anymore With the proliferation of camera deployments, comes the need to automate and enhance the ability to monitor the video streams and generate insights from them, as well as to make streaming and storage of video more efficient and cost-effective. This is where artificial intelligence (AI) comes to play. Since traditional cloud-based AI models often suffer from latency issues, not... --- - Published: 2024-11-09 - Modified: 2025-01-09 - URL: https://hailo.ai/ja/blog/ai-cameras-from-vision-to-insights/ - Categories: Edge AI Developer, Edge AI Device, Intelligent Camera - Translation Priorities: 可选 As the camera market is booming, so does the need to empower cameras with artificial intelligence (AI). In this blog post we discuss the need for on-camera AI and how it enhances video quality and empowers advanced video analytics. We will look at some typical cameras and applications and try to estimate the AI budget required to execute the different scenarios. The Essential Role of AI-Powered Cameras in Shaping a Smarter WorldAs the camera market is booming, so does the need to empower cameras with artificial intelligence (AI). In this blog post we discuss the need for on-camera AI and how it enhances video quality and empowers advanced video analytics. We will look at some typical cameras and applications and try to estimate the AI budget required to execute the different scenarios. Cameras. Cameras Everywhere In today’s tech-driven world, cameras have become an integral part of our daily lives, and we are used to constantly recording videos and being recorded. The rapid deployment of IP cameras in residential homes, commercial and public space and the industrial sector, is fueling an unprecedented growth in the market, which is estimated by ABI Research to reach 200 million cameras by 2027, with a revenue of $35B. The most significant growth driver for this market is the ability to improve safety and security through video surveillance. From enhancing home security to monitoring public safety and optimizing traffic management, smart cameras provide countless benefits and endless opportunities for a smarter, safer world. AI isn’t only cloud anymore With the proliferation of camera deployments, comes the need to automate and enhance the ability to monitor the video streams and generate insights from them, as well as to make streaming and storage of video more efficient and cost-effective. This is where artificial intelligence (AI) comes to play. Since traditional cloud-based AI models often suffer from latency issues, not... --- > Hailo AI’s autonomous parking solutions are here. Discover the futuristic parking lot technologies of automated parking AI. Learn more today! - Published: 2023-08-20 - Modified: 2025-03-20 - URL: https://hailo.ai/blog/backing-into-the-future-unlocking-the-potential-of-automated-parking/ - Categories: ADAS, Automotive - Tags: Automated parking, Driving safety - Translation Priorities: Optional The technology advancements and market drivers that accelerate the transition to automated parking Beyond Scratches: The High Toll of Parking Accidents  There’s always this annoying little blind spot when you reverse into a parking spot – the column that you missed, the wall that’s just a bit closer than you thought or the curb protrusion that’s just a little too high. It’s so frustrating to bump or scratch your car, but that’s the smallest of parking accident problems. Not many people realize the magnitude of parking accidents, and the financial and social impact they have. It is estimated that one in five vehicle accidents occur in parking lots, about a ¼ of them involve driving backwards, and a 1/3 are reported to be a result of driver’s distraction. Over 500 people die every year in parking lots and garages and 60,000 more are injured. As cities continue to expand and city centers become densely populated, the demand for parking spaces far exceeds the available supply. Residents, business people that commute daily to the city center, tourists, and other visitors are all competing over a limited number of free parking spaces. This results in frustration and stress which in turn increase the number of parking lot accidents. The wasting of valuable time, which could have been spent on more productive or enjoyable activities, and the increased traffic congestion are an additional detrimental by products of the parking space scarcity. For elderly or disabled drivers, finding an easy-to-access parking spot is not just a matter of convenience, but rather a crucial aspect of their quality... --- - Published: 2023-08-20 - Modified: 2024-08-25 - URL: https://hailo.ai/de/blog/backing-into-the-future-unlocking-the-potential-of-automated-parking/ - Categories: ADAS, Automotive - Tags: Automated parking, Driving safety - Translation Priorities: Optional The technology advancements and market drivers that accelerate the transition to automated parking Beyond Scratches: The High Toll of Parking Accidents There’s always this annoying little blind spot when you reverse into a parking spot – the column that you missed, the wall that’s just a bit closer than you thought or the curb protrusion that’s just a little too high. It’s so frustrating to bump or scratch your car, but that’s the smallest of parking accident problems. Not many people realize the magnitude of parking accidents, and the financial and social impact they have. It is estimated that one in five vehicle accidents occur in parking lots, about a ¼ of them involve driving backwards, and a 1/3 are reported to be a result of driver’s distraction. Over 500 people die every year in parking lots and garages and 60,000 more are injured. As cities continue to expand and city centers become densely populated, the demand for parking spaces far exceeds the available supply. Residents, business people that commute daily to the city center, tourists, and other visitors are all competing over a limited number of free parking spaces. This results in frustration and stress which in turn increase the number of parking lot accidents. The wasting of valuable time, which could have been spent on more productive or enjoyable activities, and the increased traffic congestion are an additional detrimental by products of the parking space scarcity. For elderly or disabled drivers, finding an easy-to-access parking spot is not just a matter of convenience, but rather a crucial aspect of their quality... --- > 由先进人工智能和传感器提供支持的自动停车正在改变我们的停车方式，减少事故并提高道路安全。 - Published: 2023-08-20 - Modified: 2024-11-06 - URL: https://hailo.ai/zh-hans/blog/backing-into-the-future-unlocking-the-potential-of-automated-parking/ - Categories: ADAS, Automotive - Tags: Automated parking, Driving safety - Translation Priorities: 可选 The technology advancements and market drivers that accelerate the transition to automated parking Beyond Scratches: The High Toll of Parking Accidents There’s always this annoying little blind spot when you reverse into a parking spot – the column that you missed, the wall that’s just a bit closer than you thought or the curb protrusion that’s just a little too high. It’s so frustrating to bump or scratch your car, but that’s the smallest of parking accident problems. Not many people realize the magnitude of parking accidents, and the financial and social impact they have. It is estimated that one in five vehicle accidents occur in parking lots, about a ¼ of them involve driving backwards, and a 1/3 are reported to be a result of driver’s distraction. Over 500 people die every year in parking lots and garages and 60,000 more are injured. As cities continue to expand and city centers become densely populated, the demand for parking spaces far exceeds the available supply. Residents, business people that commute daily to the city center, tourists, and other visitors are all competing over a limited number of free parking spaces. This results in frustration and stress which in turn increase the number of parking lot accidents. The wasting of valuable time, which could have been spent on more productive or enjoyable activities, and the increased traffic congestion are an additional detrimental by products of the parking space scarcity. For elderly or disabled drivers, finding an easy-to-access parking spot is not just a matter of convenience, but rather a crucial aspect of their quality... --- - Published: 2023-08-20 - Modified: 2024-12-15 - URL: https://hailo.ai/ja/blog/backing-into-the-future-unlocking-the-potential-of-automated-parking/ - Categories: ADAS, Automotive - Tags: Automated parking, Driving safety - Translation Priorities: 可选 The technology advancements and market drivers that accelerate the transition to automated parking Beyond Scratches: The High Toll of Parking Accidents There’s always this annoying little blind spot when you reverse into a parking spot – the column that you missed, the wall that’s just a bit closer than you thought or the curb protrusion that’s just a little too high. It’s so frustrating to bump or scratch your car, but that’s the smallest of parking accident problems. Not many people realize the magnitude of parking accidents, and the financial and social impact they have. It is estimated that one in five vehicle accidents occur in parking lots, about a ¼ of them involve driving backwards, and a 1/3 are reported to be a result of driver’s distraction. Over 500 people die every year in parking lots and garages and 60,000 more are injured. As cities continue to expand and city centers become densely populated, the demand for parking spaces far exceeds the available supply. Residents, business people that commute daily to the city center, tourists, and other visitors are all competing over a limited number of free parking spaces. This results in frustration and stress which in turn increase the number of parking lot accidents. The wasting of valuable time, which could have been spent on more productive or enjoyable activities, and the increased traffic congestion are an additional detrimental by products of the parking space scarcity. For elderly or disabled drivers, finding an easy-to-access parking spot is not just a matter of convenience, but rather a crucial aspect of their quality... --- > Harnessing the power of ADAS development to deliver state-of-the-art AI driving solutions. Discover how we are paving the way for safer and smarter roads. - Published: 2023-07-24 - Modified: 2024-08-25 - URL: https://hailo.ai/blog/leveraging-vendor-partnerships-for-adas-success-leddartech-and-hailo/ - Categories: ADAS, AI Software, Automotive - Translation Priorities: Optional It’s a late summer evening, you’ve had a long day at work and all you want to do is get home and relax, but the usual horrible traffic jam is worrying you. The thought of spending the next thirty minutes switching between the accelerator and brake pedals is frustrating. As per the INRIX 2022 report, the average driver in London, Chicago and Paris lost respectively 156, 155 and 138 hours of the year in traffic jams, other cities showing the same patterns. ADAS developers realize that developing a solution to combat such situations is not just a matter of convenience but also safety, mental health, and overall mobility experience. ADAS features such as traffic jam assist, highway assist and automatic lane change improve the driver’s experience. However, as highlighted in JD Power’s 2022 Mobility Confidence Index Study, consumer understanding of ADAS is not yet ubiquitous, and inconsistencies in operation and performance could worsen consumer understanding and acceptance. Improved ADAS performance, enabled by better sensor fusion and perception, will likely increase ADAS adoption, trust, and confidence. A goal that all ADAS developers work towards. Unlocking ADAS: Challenges Faced by ADAS Developers Automotive OEMs and suppliers are in a fierce race to deliver the latest ADAS features quickly due to the significant positive financial impact of ADAS on new vehicle sales, market share and profitability. ADAS developers must juggle multiple demands when developing ADAS features, such as: Performance and innovation Cost optimization Ease of integration Scalability across vehicle models and platforms Flexibility... --- - Published: 2023-07-24 - Modified: 2024-08-25 - URL: https://hailo.ai/zh-hans/blog/leveraging-vendor-partnerships-for-adas-success-leddartech-and-hailo/ - Categories: ADAS, AI Software, Automotive - Translation Priorities: 可选 It’s a late summer evening, you’ve had a long day at work and all you want to do is get home and relax, but the usual horrible traffic jam is worrying you. The thought of spending the next thirty minutes switching between the accelerator and brake pedals is frustrating. As per the INRIX 2022 report, the average driver in London, Chicago and Paris lost respectively 156, 155 and 138 hours of the year in traffic jams, other cities showing the same patterns. ADAS developers realize that developing a solution to combat such situations is not just a matter of convenience but also safety, mental health, and overall mobility experience. ADAS features such as traffic jam assist, highway assist and automatic lane change improve the driver’s experience. However, as highlighted in JD Power’s 2022 Mobility Confidence Index Study, consumer understanding of ADAS is not yet ubiquitous, and inconsistencies in operation and performance could worsen consumer understanding and acceptance. Improved ADAS performance, enabled by better sensor fusion and perception, will likely increase ADAS adoption, trust, and confidence. A goal that all ADAS developers work towards. Unlocking ADAS: Challenges Faced by ADAS Developers Automotive OEMs and suppliers are in a fierce race to deliver the latest ADAS features quickly due to the significant positive financial impact of ADAS on new vehicle sales, market share and profitability. ADAS developers must juggle multiple demands when developing ADAS features, such as: Performance and innovation Cost optimization Ease of integration Scalability across vehicle models and platforms Flexibility... --- > Nutzen Sie die Leistungsfähigkeit der ADAS-Entwicklung, um hochmoderne KI-Fahrlösungen bereitzustellen. Für sicherere und intelligentere Straßen. - Published: 2023-07-24 - Modified: 2024-08-25 - URL: https://hailo.ai/de/blog/leveraging-vendor-partnerships-for-adas-success-leddartech-and-hailo/ - Categories: ADAS, ADAS, AI Software, Automotive, Automotive - Translation Priorities: Optional It’s a late summer evening, you’ve had a long day at work and all you want to do is get home and relax, but the usual horrible traffic jam is worrying you. The thought of spending the next thirty minutes switching between the accelerator and brake pedals is frustrating. As per the INRIX 2022 report, the average driver in London, Chicago and Paris lost respectively 156, 155 and 138 hours of the year in traffic jams, other cities showing the same patterns. ADAS developers realize that developing a solution to combat such situations is not just a matter of convenience but also safety, mental health, and overall mobility experience. ADAS features such as traffic jam assist, highway assist and automatic lane change improve the driver’s experience. However, as highlighted in JD Power’s 2022 Mobility Confidence Index Study, consumer understanding of ADAS is not yet ubiquitous, and inconsistencies in operation and performance could worsen consumer understanding and acceptance. Improved ADAS performance, enabled by better sensor fusion and perception, will likely increase ADAS adoption, trust, and confidence. A goal that all ADAS developers work towards. Unlocking ADAS: Challenges Faced by ADAS Developers Automotive OEMs and suppliers are in a fierce race to deliver the latest ADAS features quickly due to the significant positive financial impact of ADAS on new vehicle sales, market share and profitability. ADAS developers must juggle multiple demands when developing ADAS features, such as: Performance and innovation Cost optimization Ease of integration Scalability across vehicle models and platforms Flexibility... --- - Published: 2023-07-24 - Modified: 2024-09-09 - URL: https://hailo.ai/ja/blog/leveraging-vendor-partnerships-for-adas-success-leddartech-and-hailo/ - Categories: ADAS, AI Software, Automotive - Translation Priorities: 可选 It’s a late summer evening, you’ve had a long day at work and all you want to do is get home and relax, but the usual horrible traffic jam is worrying you. The thought of spending the next thirty minutes switching between the accelerator and brake pedals is frustrating. As per the INRIX 2022 report, the average driver in London, Chicago and Paris lost respectively 156, 155 and 138 hours of the year in traffic jams, other cities showing the same patterns. ADAS developers realize that developing a solution to combat such situations is not just a matter of convenience but also safety, mental health, and overall mobility experience. ADAS features such as traffic jam assist, highway assist and automatic lane change improve the driver’s experience. However, as highlighted in JD Power’s 2022 Mobility Confidence Index Study, consumer understanding of ADAS is not yet ubiquitous, and inconsistencies in operation and performance could worsen consumer understanding and acceptance. Improved ADAS performance, enabled by better sensor fusion and perception, will likely increase ADAS adoption, trust, and confidence. A goal that all ADAS developers work towards. Unlocking ADAS: Challenges Faced by ADAS Developers Automotive OEMs and suppliers are in a fierce race to deliver the latest ADAS features quickly due to the significant positive financial impact of ADAS on new vehicle sales, market share and profitability. ADAS developers must juggle multiple demands when developing ADAS features, such as: Performance and innovation Cost optimization Ease of integration Scalability across vehicle models and platforms Flexibility... --- > Advanced AI object detection for edge devices with Hailo. Learn how to make the right choices for your projects. Discover AI solutions today! - Published: 2022-10-06 - Modified: 2025-05-05 - URL: https://hailo.ai/blog/ai-object-detection-on-the-edge-making-the-right-choice/ - Categories: Edge AI Developer, Edge AI Device, Object Detection - Translation Priorities: Optional When choosing an AI object detection network for edge devices, there are many factors you should consider: compute power, memory resources, and many more. Which ones? This blogpost outlines everything you need to know when choosing an object detection network for your edge application. But remember, the Hailo-8 processor provides high-performance computing on the edge and can have a prominent role in improving the accuracy of the network.   Object detection in computer vision classifies and localizes all the objects in an image. It is widely used in Automotive, Smart City, Smart Home, and Industry 4. 0 applications, among others.   However, running AI object detection on the edge has some drawbacks as well. One reason is that compute and memory are limited on edge devices, which limits the choice of the object detection network. For example, the standard mobile/CPU regime defined in the deep learning literature usually allows approximately 800M FLOPS per frame. Under this regime, the chosen object detector would need to be highly efficient and small, for example, MobileNet-v2-SSD (760M FLOPS for ~0. 1MPixel input). Another set of constraints in edge devices is power consumption and heat dissipation, which also limits the processing throughput.   Figure 1 - system design for an AI-enabled sensor at the edge However, running object detection on the edge has some drawbacks as well. One reason is that compute and memory are limited on edge devices, which limits the choice of the object detection network. For example, the standard mobile/CPU regime defined in the deep learning literature usually allows approximately... --- > Objekterkennungs-KI am Edge: Alles, was Sie wissen müssen, wenn Sie ein Objekterkennungsnetzwerk für Ihre Edge-Anwendung auswählen. - Published: 2022-10-06 - Modified: 2024-08-25 - URL: https://hailo.ai/de/blog/object-detection-at-the-edge-making-the-right-choice/ - Categories: Edge AI Developer, Edge AI Developer, Edge AI Device, Edge AI Device, Object Detection - Translation Priorities: Optional When choosing an AI object detection network for edge devices, there are many factors you should consider: compute power, memory resources, and many more. Which ones? This blogpost outlines everything you need to know when choosing an object detection network for your edge application. But remember, the Hailo-8 processor provides high-performance computing on the edge and can have a prominent role in improving the accuracy of the network.   Object detection in computer vision classifies and localizes all the objects in an image. It is widely used in Automotive, Smart City, Smart Home, and Industry 4. 0 applications, among others.   However, running AI object detection on the edge has some drawbacks as well. One reason is that compute and memory are limited on edge devices, which limits the choice of the object detection network. For example, the standard mobile/CPU regime defined in the deep learning literature usually allows approximately 800M FLOPS per frame. Under this regime, the chosen object detector would need to be highly efficient and small, for example, MobileNet-v2-SSD (760M FLOPS for ~0. 1MPixel input). Another set of constraints in edge devices is power consumption and heat dissipation, which also limits the processing throughput.   Figure 1 - system design for an AI-enabled sensor at the edge However, running object detection on the edge has some drawbacks as well. One reason is that compute and memory are limited on edge devices, which limits the choice of the object detection network. For example, the standard mobile/CPU regime defined in the deep learning literature usually allows approximately... --- - Published: 2022-10-06 - Modified: 2025-01-09 - URL: https://hailo.ai/zh-hans/blog/ai-object-detection-on-the-edge-making-the-right-choice/ - Categories: Edge AI Developer, Edge AI Device, Object Detection - Translation Priorities: 可选 When choosing an AI object detection network for edge devices, there are many factors you should consider: compute power, memory resources, and many more. Which ones? This blogpost outlines everything you need to know when choosing an object detection network for your edge application. But remember, the Hailo-8 processor provides high-performance computing on the edge and can have a prominent role in improving the accuracy of the network.   Object detection in computer vision classifies and localizes all the objects in an image. It is widely used in Automotive, Smart City, Smart Home, and Industry 4. 0 applications, among others. However, running AI object detection on the edge has some drawbacks as well. One reason is that compute and memory are limited on edge devices, which limits the choice of the object detection network. For example, the standard mobile/CPU regime defined in the deep learning literature usually allows approximately 800M FLOPS per frame. Under this regime, the chosen object detector would need to be highly efficient and small, for example, MobileNet-v2-SSD (760M FLOPS for ~0. 1MPixel input). Another set of constraints in edge devices is power consumption and heat dissipation, which also limits the processing throughput.   Figure 1 - system design for an AI-enabled sensor at the edge However, running object detection on the edge has some drawbacks as well. One reason is that compute and memory are limited on edge devices, which limits the choice of the object detection network. For example, the standard mobile/CPU regime defined in the deep learning literature usually allows approximately 800M... --- - Published: 2022-10-06 - Modified: 2025-01-09 - URL: https://hailo.ai/ja/blog/ai-object-detection-on-the-edge-making-the-right-choice/ - Categories: Edge AI Developer, Edge AI Device, Object Detection - Translation Priorities: 可选 When choosing an AI object detection network for edge devices, there are many factors you should consider: compute power, memory resources, and many more. Which ones? This blogpost outlines everything you need to know when choosing an object detection network for your edge application. But remember, the Hailo-8 processor provides high-performance computing on the edge and can have a prominent role in improving the accuracy of the network.   Object detection in computer vision classifies and localizes all the objects in an image. It is widely used in Automotive, Smart City, Smart Home, and Industry 4. 0 applications, among others. However, running AI object detection on the edge has some drawbacks as well. One reason is that compute and memory are limited on edge devices, which limits the choice of the object detection network. For example, the standard mobile/CPU regime defined in the deep learning literature usually allows approximately 800M FLOPS per frame. Under this regime, the chosen object detector would need to be highly efficient and small, for example, MobileNet-v2-SSD (760M FLOPS for ~0. 1MPixel input). Another set of constraints in edge devices is power consumption and heat dissipation, which also limits the processing throughput.   Figure 1 - system design for an AI-enabled sensor at the edge However, running object detection on the edge has some drawbacks as well. One reason is that compute and memory are limited on edge devices, which limits the choice of the object detection network. For example, the standard mobile/CPU regime defined in the deep learning literature usually allows approximately 800M... --- > See how multi-camera tracking & multi-person tracking with Hailo-8 boosts operations for retail and security efficient accurate insights. - Published: 2022-08-07 - Modified: 2025-04-20 - URL: https://hailo.ai/blog/multi-camera-multi-person-re-identification/ - Categories: Edge AI Box, Edge AI Device, Security, Smart Retail, Surveillance - Translation Priorities: Optional Multi-person re-identification across different streams is essential for security and retail applications. Multi-person re-identification across different streams is essential for security and retail applications, facilitated by multi-camera tracking. This includes the identification of a specific person multiple times, either in a specific location over time, or along a trail between multiple locations. High performance computing is required to achieve this. Hailo-8 AI processor delivers the efficiency needed for accurate real-time multi-person re-identification on edge devices while enhancing video analytics and keeping costs down without compromising people’s privacy. In this blog post, we present an end-to-end reference pipeline to perform this task with precision using multi-camera tracking technology.   https://youtu. be/Gos90gTxaWw Multi-person re-identification tracking on multiple video streams is a common feature in video surveillance systems. The goal of this feature, which is implemented usually by deep learning, is to detect and identify people across different streams and throughout the video. The multi-person re-identification feature is used for safety, security, and data analysis, which reveals valuable information on customers, visitors, and employees’ behavior. Although it is generally considered a widely used technology, occlusions and the different conditions of each camera (such as perspective and illumination), generate a major challenge for accurate tracking. Figure 1 - Multi-Camera tracking for multi-Person re-identification output. The system can do multi-person re-identification across different frames and different views.   Advantages of running a multi-camera tracking with multi-person re-identification application on edge devices include: Preserve privacy and improve data protection by eliminating the need to send raw video.   Improve the detection latency which is crucial for real-time alerts.... --- > Mehrkamera-Tracking und Mehrpersonen-Tracking mit Hailo-8 verbessern den Einzelhandelsbetrieb und liefern präzise Einblicke in die Sicherheit. - Published: 2022-08-07 - Modified: 2024-12-24 - URL: https://hailo.ai/de/blog/multi-camera-multi-person-re-identification/ - Categories: Edge AI Box, Edge AI Device, Edge AI Device, Security, Security, Smart Retail, Surveillance, Surveillance - Translation Priorities: Optional Multi-person re-identification across different streams is essential for security and retail applications. Die Wiedererkennung mehrerer Personen über verschiedene Datenströme hinweg ist für Sicherheits- und Einzelhandelsanwendungen unerlässlich, und wird durch die Verfolgung mit mehreren Kameras erleichtert. Dies beinhaltet die mehrfache Erkennung einer bestimmten Person, entweder an einem bestimmten Ort im Laufe der Zeit oder entlang einer Spur zwischen mehreren Orten.  Umdies zu erreichen, ist Hochleistungsrechnen erforderlich.  Der Hailo-8 KI-Prozessor bietet die Effizienz, die für einegenaue Wiedererkennung mehrerer Personen in Echtzeit auf Randgeräte erforderlich ist. Gleichzeitig verbessert er die Videoanalyse und senkt die Kosten, ohne die Privatsphäre der Benutzer zu gefährden. In diesem Blogbeitrag stellen wir eine umfassende Referenzpipeline vor, mit der dieseAufgabe mithilfe der Multikamera-Tracking-Technologie präzise ausgeführt werdenkann.   https://youtu. be/Gos90gTxaWw Die Verfolgung der Wiederidentifikation mehrerer Personen auf mehreren Videostreams ist ein übliches Merkmal von Videoüberwachungssystemen. Das Ziel dieser Funktion, die in der Regel durch Deep Learning durchgeführt wird, besteht darin, Personen in verschiedenen Streams und im gesamten Video zu erkennen und zu identifizieren. Die Funktion zur erneuten Identifizierung mehrerer Personen wird für Sicherheits- und Datenanalysen verwendet, wodurch wertvolle Informationen über das Verhalten von Kunden, Besuchern und Mitarbeitern erhalten werden. Obwohl sie allgemein als eine weit verbreitete Technologie angesehen wird, stellen Verdeckungen und die unterschiedlichen Bedingungen der einzelnen Kameras (z. B. Perspektive und Beleuchtung) eine große Herausforderung für die genaue Verfolgung dar. Abbildung 1 — Multikamera-Tracking für die Ausgabe der Wiedererkennung mehrerer Personen. Das System kann mehrere Personen über verschiedene Bilder und Ansichten hinweg wiedererkennen.   Zu den Vorteilen der Ausführung einer Multikamera-Tracking-Anwendung mit Wiedererkennung mehrerer Personen auf Randgeräte gehören: Wahrung der Privatsphäre... --- > 了解 Hailo-8 的多摄像机跟踪和多人跟踪如何促进零售和安全运营的高效准确洞察。 - Published: 2022-08-07 - Modified: 2024-12-24 - URL: https://hailo.ai/zh-hans/blog/multi-camera-multi-person-re-identification/ - Categories: Edge AI Box, Edge AI Device, Security, Smart Retail, Surveillance - Translation Priorities: 可选 Multi-person re-identification across different streams is essential for security and retail applications. 借助多摄像头跟踪技术，对不同数据流进行多人重识别对于安全和零售应用至关重要。这包括多次识别特定人员，要么在一段时间内在特定的位置进行识别，要么沿着多个地点之间的路径进行识别。要实现这一点，需要高性能计算。Hailo-8 AI处理器提供在边缘设备进行准确的实时多人重识别所需的效率，同时在不损害人们隐私的情况下增强视频分析并降低成本。在这篇博客文章中，我们提供了端到端的参考管道，使用多摄像头跟踪技术精确地执行这项任务。 https://youtu. be/Gos90gTxaWw 对多个视频流进行多人重识别跟踪是视频监控系统的常见功能。此功能通常通过深度学习实现，其目标是在不同流和整个视频中检测和识别人员。多人重识别功能被用于安全、安保和数据分析，揭示有关客户、访客和员工行为的有价值信息。尽管它通常被认为是一种广泛使用的技术，但遮挡和各摄像机的不同条件（例如视角和照明）对精确跟踪造成了重大挑战。 图1 — 面向多人重识别输出的多摄像头跟踪技术。该系统可以在不同的画面和不同的视图上进行多人重识别。  在边缘设备上使用多人重识别应用进行多摄像头跟踪的优势包括 无需发送原始视频，从而保护隐私并改善数据保护。  改善检测延迟，这对于实时警报至关重要. Hailo-8是理想的AI加速器，可在边缘设备上进行准确的实时人员重识别。它的计算能力还能够同时对许多人进行高精度的处理，这对于实现高质量的重识别至关重要。通过安装和维护单个AI加速器来实时处理多个摄像头，Hailo-8还降低了系统成本。 Hailo TAPPAS多摄像头重识别和跟踪管道是通过在嵌入式主机上使用GStreamer来实现的，Hailo-8实时运行（无需批处理），采用四个全高清输入分辨率的RTSP IP摄像头。主机通过以太网获取编码后的视频并对其进行解码。解码后的帧通过PCIe发送到Hailo-8上进行处理，最终输出显示在屏幕上。 图2 — 采用Hailo-8的多摄像头多人应用的系统图纸。  应用管道 下图描绘了整个应用管道。首先，对编码后的输入视频进行解码和去扭曲，以获得对齐的帧，以便进行处理。去扭曲是一种常见的计算机视觉组件，用于消除由摄像头引起的任何失真，例如，鱼眼失真在监控摄像头中很常见。接下来，这些帧被发送到Hailo-8 AI处理器，处理器可以检测每帧中的所有人和面孔。我们使用Hailo GStreamer跟踪器对每个视频流中的物体进行初始跟踪。最后，将每个人从原始画面中裁剪出来，然后输入到重识别（Re-ID）网络中。此网络输出一个代表每个人的嵌入向量，该向量可以在不同的摄像头之间进行比较。嵌入向量存储在一个名为“图库”（gallery）的数据库中，通过在其中进行搜索，我们为每个人分配了最终的ID。最终输出还包括一个人脸匿名块，该块可以模糊每张面孔，以保护画面中人们的隐私。  所有的神经网络（NN）模型都是使用Hailo Dataflow Compiler进行编译，预训练的权重和预编译的模型在Hailo Model Zoo中发布。Hailo Model Zoo还为自定义数据集提供重新训练docker环境，以便适应其他场景。我们注意到，所有模型都是在相对通用的用例中进行训练的，可以针对特定场景进行优化（在大小/精度/帧率方面）。  图3 — Hailo-8多摄像头多人重识别应用的计算机视觉管道。蓝色 - 在Hailo-8设备上运行的区块，橙色 - 在嵌入式主机上运行的区块. 多人/人脸检测 人/人脸检测网络基于YOLOv5s，分为两类：人和人脸。YoloV5是一款2020年发布的精确的单级物体探测器，使用Pytorch进行训练。为了训练探测网络，我们整理了几个不同的数据集，并将它们调整为相同的注释格式。请注意，公共数据集，例如COCO、Open Images等，可能仅包含人物或人脸注释，为了使用它们，我们为这两个类别都生成了完整的注释。对于人脸注释，我们使用了在公开数据集上训练的先进的人脸检测模型。使用强大的神经网络（例如YoloV5）来检测人和人脸意味着我们能够高精度和远距离检测到他们；因此，让应用能够完成检测和跟踪即，使是很小的物体也不例外 人重识别 人重识别（Re-ID）网络基于Rep-VGG-A0，每次查询输出长度为2048的单个嵌入向量。该网络是使用以下存储库在Pytorch中训练的。为了提高验证数据集（Market-1501）的Rank-1准确性，我们将不同的重识别数据集合并到一个训练程序中。使用更大、更多样化的训练数据（来自多个来源）帮助我们生成了一个更强大的网络，可以更好地推广到现实场景。在Hailo Model Zoo中，我们提供重新训练说明和完整的docker环境，以便使用我们预训练的权重来训练网络。 图4 — 重识别网络示例。该网络为每个人输出一个向量，因此我们可以将其与人群库进行比较，以获得最终的验证准确性 使用Hailo TAPPAS来部署管道 我们已经将该应用作为Hailo TAPPAS的一部分进行发布。该示例应用使用GStreamer C++中构建管道，允许您使用视频文件或RTSP摄像头。允许您控制应用的其他参数包括设置探测器（例如检测阈值）、跟踪器（例如，保持/丢失帧率）和质量估计（最低质量阈值）的参数。  Hailo Model Zoo还允许您使用自己的数据重新训练 神经网络，并将它们移植到TAPPAS应用中，以便快速适应领域和进行自定义。多摄像头多人重识别应用程序的目标是为在Hailo-8和嵌入式主机处理器上建立监控管道提供快速的原型设计和坚实的基准。 作为HailoRT（Hailo运行时库）的一部分，我们发布了一个用于在Hailo-8芯片（libgsthailo）上进行推理的GStreamer插件。该插件负责芯片上的整个配置和推理过程，这使得Hailo-8可以轻松直接地集成到您的GStreamer管道中。它还可以在单个Hailo-8芯片上进行多网络管道推理，以简化复杂的管道。我们引入的另一个HaioRT组件是网络调度器。这个HailoRT组件通过自动网络切换来简化在单个Hailo设备上运行多个网络的流程。网络调度器不会手动决定哪个网络的运行时间，而是自动控制每个网络的运行时间。使用调度器让使用Hailo-8进行管道开发变得更简洁、更简单、更高效。 除了HailoRT外，我们在这个应用中还引入了以下GStreamer插件：  去扭曲：在TAPPAS中实施的这个GStreamer插件允许您修复摄像头失真。去扭曲是使用OpenCV实现的，目前正在修复鱼眼失真。  方框匿名化：在TAPPAS中实施的这个GStreamer插件允许您在给定预测方框的情况下模糊图像中的方框。例如，在预测所有人脸之后，对图像进行人脸匿名化。  图库搜索：这个GStreamer插件可向管道中添加数据库组件。图库组件允许您添加新对象并在数据库中搜索匹配项。在本应用中，我们将推送向量并将其与新向量进行比较，以关联不同摄像头和时间戳之间的预测。 我们可以看到，在采用6个摄像头和投入500美元预算的小型企业中，RSC101可以有效地支持所有必需的监控功能。实际上，RSC101的性能超出了要求。如果利用集成的高算力AI处理器进行所有视频分析，就可以使用先进的深度学习算法，从而实现高性能和高级功能。26TOPS的AI算力为未来的增强提供了保障，支持向先进的深度学习算法迁移，以检测更多的事件类型。请注意，RSC101的功能使其能够作为极为经济实惠的解决方案。在视频分析算力得到充分利用的情况下，它为现有监控系统增加了对额外4-8个摄像头的支持。 性能 下表汇总了在Hailo-8和x86主机处理器上使用四个全高清输入分辨率（1920×1080）的RTSP摄像头的多摄像头多人跟踪应用的性能，以及神经网络独立性能的细分信息。  Hailo多摄像头多人重识别应用提供了部署在GStreamer中的完整参考管道，其中包含Hailo TAPPAS和针对每个神经网络的再培训功能，以便使用Hailo Model Zoo进行自定义。此应用为您使用Hailo-8构建VMS产品提供了基准。如需了解更多信息，请访问我们的TAPPAS文档。  本文由Tamir Tapuhi、Amit Klinger、Omer Sholev、Rotem Bar和Yuval Belzer合作撰写. --- > Hailo-8 によるマルチカメラ追跡と複数人物追跡により、小売業やセキュリティ業務が効率化され、正確な洞察が得られる仕組みをご覧ください。 - Published: 2022-08-07 - Modified: 2024-12-24 - URL: https://hailo.ai/ja/blog/multi-camera-multi-person-re-identification/ - Categories: Edge AI Box, Edge AI Device, Security, Smart Retail, Surveillance - Translation Priorities: 可选 Multi-person re-identification across different streams is essential for security and retail applications. 異なるストリームにわたる複数人物の再同定は、マルチカメラによる追跡によって実現されるセキュリティや小売アプリケーションに不可欠です。これには、特定の場所で、時間をかけて特定の人物を複数回同定したり、複数の場所をつなぐ痕跡に沿って同定したりすることが含まれます。これを実現するには、ハイパフォーマンス・コンピューティングが必要です。Hailo-8 AIプロセッサは、エッジデバイス上で正確なリアルタイムの複数人物再同定に必要な効率性を提供すると同時に、ビデオ解析を強化し、人々のプライバシーを損なうことなくコストを抑えます。このブログ記事では、マルチカメラトラッキング技術を使用してこのタスクを正確に実行するためのエンドツーエンドのリファレンス・パイプラインをご紹介します。  https://youtu. be/Gos90gTxaWw 複数のビデオストリームでの複数人による再識別追跡は、ビデオ監視システムにおける共通機能です。この機能は通常、ディープラーニングによって実装され、さまざまなストリームやビデオ全体にわたって人物を検出し、識別することを目的としています。複数人物の再識別機能は、安全、セキュリティ、およびデータ分析に使用され、顧客、訪問者、従業員の行動に関する貴重な情報を明らかにします。一般的に広く使用されているテクノロジーと考えられていますが、遮蔽や各カメラのさまざまな条件（画角や照明など）により、正確なトラッキングには大きな課題が生じます。 図 1 — 複数人物再同定出力のためのマルチカメラトラッキング。このシステムは、異なるフレームやビューを横断する複数の人物を再同定できます。  エッジデバイス上で複数人物再同定アプリケーションを使用してマルチカメラトラッキングを実行する利点は次のとおりです: 未加工の動画を送信する必要がなくなるため、プライバシーを保護し、データ保護を向上します。  リアルタイムアラートに不可欠な検出遅延を改善します。  Hailo-8は、エッジデバイスで正確なリアルタイムの人物再同定を実現するのに最適なAIアクセラレータです。また、その計算能力は、高品質の再同定に不可欠な高い確度で、多数の人物を同時に高精度で処理することを可能にします。Hailo-8は、1台の AIアクセラレータ で複数のカメラをリアルタイムに処理できるため、設置やメンテナンスにかかるシステムコストを削減します。 Hailo TAPPAS マルチカメラ再同定およびトラッキング・パイプラインは、組み込みホスト上のGStreamerと、FHD入力解像度の4台のRTSP IPカメラでHailo-8をリアルタイムで実行（バッチ処理なし）することで実装されています。ホストはエンコードされたビデオをイーサネット経由で取得し、デコードします。デコードされたフレームは PCIe 経由で Hailo-8 に送られ処理され、最終出力が画面に表示されます。 図2 — Hailo-8で動作するマルチカメラ複数人物用アプリケーションのシステム図. アプリケーション・パイプライン アプリケーション・パイプライン全体を次の図に示します。まず、エンコードされた入力がデコードされ、デワープされて、処理のために整列されたフレームが取得されます。デワーピングは、カメラによって生じる歪みを除去するために使用される一般的なコンピュータ・ビジョン・コンポーネントです。たとえば、魚眼歪みは監視カメラでよく見られます。次に、フレームはHailo-8AIプロセッサに送信され、各フレーム内のすべての人物と顔が検出されます。各ストリーム内のオブジェクトの初期トラッキングには、Hailo GStreamer トラッカーを使用します。最後に、各人物は元のフレームから切り取られ、再同定ネットワークに入力されます。このネットワークは、各人物を表す埋め込みベクトルを出力し、異なるカメラ間で比較することができます。埋め込みは「ギャラリー」と呼ばれるデータベースに保存され、その中を検索して、各人に最終的なIDを割り当てます。最終的な出力には、フレーム内の人物のプライバシーを保護するために、各々の顔をぼかす顔匿名化ブロックも含まれています。  すべてのニューラルネットワーク（NN）モデルは、Hailo データフローコンパイラ を使用してコンパイルされ、事前学習済みのウェイトとコンパイル済みモデルは Hailo Model Zooでリリースされました。Hailo Model Zooには、他のシナリオへの適応が容易になるように、カスタムデータセット用の再学習ドッカー環境も提供します。すべてのモデルは比較的一般的なユースケースで学習されており、特定のシナリオに合わせて（サイズ/精度/fpsの点で）最適化できることに注意してください。 図3 — Hailo-8によるマルチカメラ複数人物再同定アプリケーションのCVパイプライン。青色は Hailo-8 デバイス上で実行されるブロック、オレンジ色は組み込みホスト上で実行されるブロック。  複数人/顔検出 人物/顔検出ネットワークはYOLOv5sをベースにしており、人物と顔の2つのクラスがあります。YOLOv5は、2020年にリリースされた高精度な 1段階の物体検出器であり、Pytorchで学習されています。検出ネットワークを学習するために、いくつかの異なるデータセットをキュレートし、それらを同じアノテーションフォーマットに揃えました。COCO、Open Imagesなどの公開データセットには、人物または顔のアノテーションしか含まれていない場合があり、それらを使用するために両方のクラスに完全な注釈を生成したことにご留意ください。顔へのアノテーションには、公開されているデータセットで学習された最先端の顔検出モデルを使用しました。YOLOv5などの強力なNNを使用して人物や顔を検出すると、高精度かつ遠距離で検出できるため、小さな物体でも検出および追跡できます。 人物再同定 人物再同定ネットワークはrep-VGG-a0に基づいており、クエリごとに長さ 2048 の単一の埋め込みベクトルを出力します。ネットワークは、以下のリポジトリを使用してPytorchで学習されました。検証データセット（Market-1501）のRank-1の精度を向上させるために、さまざまな再同定データセットを1つの学習手順に統合しました。より大規模で多様な（複数のソースからの）学習データを使用することで、現実世界のシナリオをより適切に一般化できる堅牢なネットワークを構築することができました。Hailo Model Zooでは、再学習の手順と、事前学習したウェイトからネットワークを学習するための完全なドッカー環境を提供しています。  図 4 — 再同定ネットワークの例。ネットワークは各人物のベクトルを出力するので、それを人物のギャラリーと比較して、最終的な検証精度を得ることができます。  Hailo TAPPASを使ったパイプラインの展開 Hailo TAPPASの一部としてアプリケーションをリリースしました。サンプルアプリケーションは C++ の GStreamer を使用してパイプラインを構築し、ビデオファイルまたは RTSP カメラから実行できるようにします。アプリケーションを制御できるその他の引数には、検出器(検出閾値など)、トラッカー (キープ/ロストフレームレートなど)、品質推定 (最低品質閾値) のパラメータ設定などがあります。  Hailo Model Zooでは、独自のデータでNNを再学習し、TAPPASアプリケーションに移植することで、迅速なドメイン適応とカスタマイズを行うこともできます。マルチカメラによる複数人物再同定アプリケーションの目標は、Hailo-8と組み込みホストプロセッサ上で監視パイプラインを構築するための迅速なプロトタイピングと強固なベースラインを提供することです。  HailoRT（Hailoのランタイム・ライブラリ）の一部として、Hailo-8チップ上で推論用のGStreamerプラグイン（libgsthailo）をリリースしました。このプラグインは、チップ上の設定と推論プロセス全体を処理するため、Hailo-8をGStreamerパイプラインに簡単かつシンプルに統合できます。また、複雑なパイプラインを容易にするために、単一のHailo-8チップ上でマルチネットワーク・パイプラインの推論を可能にします。当社が導入したもう 1 つの HailoRT コンポーネントはネットワーク・スケジューラです。このHailoRTコンポーネントは、ネットワークの切り替えを自動化することで、単一のHailoデバイス上で複数のネットワークを実行する使い方を簡素化します。どのネットワークをいつ実行するかを手動で決定する代わりに、ネットワーク・スケジューラは各ネットワークの実行時間を自動的に制御します。スケジューラを使用することで、Hailo-8を使ったパイプラインの開発は、より簡潔にシンプルで効率的なものになります。  HailoRTとは別に、このアプリケーションでは以下のGStreamerプラグインも導入しました。  デワーピング: TAPPASに実装されたGStreamerプラグインを使用すると、カメラの歪みを修正できます。デワーピングはOpenCVを使用して実装されており、現在、魚眼の歪みを修正しています。  ボックスの匿名化: TAPPASに実装されたGStreamerプラグインを使用すると、予測されたボックスに含まれた画像内のボックスをぼかすことができます。たとえば、すべての顔を予測した後に、画像内の顔の匿名化を行います。  ギャラリー検索: GStreamer プラグインは、パイプラインにデータベース・コンポーネントを追加します 。ギャラリー・コンポーネントにより、新しいオブジェクトを追加し、データベース内で一致するオブジェクトを検索できます。このアプリケーションでは、再同定ベクトルをプッシュして新しいベクトルと比較し、異なるカメラやタイムスタンプ間の予測を関連付けます。 このように、RSC101は6台のカメラと500ドルの予算で、中小企業に必要なすべての監視機能を効果的にサポートします。実際、RSC101の性能は必要要件を上回っています。すべてのビデオ分析に統合された高い計算能力を持つAIプロセッサを活用することで、最先端の深層学習アルゴリズムを使用することができ、その結果、高性能で高度な機能を実現できました。26TOPS の AI 計算能力により、さまざまなイベント検出タイプの最先端の深層学習アルゴリズムへの移行をサポートすることで、将来のエンハンス開発を保証します。RSC101の機能により、最高の費用対効果のソリューションとして使用することができます。既存の監視システムに4～8台のカメラを追加することで、ビデオ解析用の計算能力がすでにフル活用されている場合にも対応できます。 性能 次の表は、FHD入力解像度（1920×1080）の4台のRTSPカメラを使用した、Hailo-8およびx86ホストプロセッサ上のマルチカメラ複数人物トラッキング・アプリケーションの性能と、NNスタンドアロン性能の内訳をまとめたものです。 Hailo マルチカメラ複数人物再同定アプリケーションは、Hailo TAPPASでGStreamerに展開されたリファレンス・パイプライン全体と、Hailo Model Zooによるカスタマイズを可能にする各NNの再学習機能を提供します。このアプリケーションは、Hailo-8を使用してVMS製品を構築するためのベースラインを提供します。詳細については、TAPPAS documentationをご覧ください。  この記事は、タミール・タプヒ、アミット・クリンガー、オマー・ショレフ、ロテム・バー、ユヴァル・ベルザーの共著です。  --- > AI video analytics are a cost-effective solution enabling real-time accurate event detection for small businesses and smart retail. - Published: 2022-06-16 - Modified: 2024-11-05 - URL: https://hailo.ai/blog/advanced-video-analytics-for-small-businesses/ - Categories: Edge AI Box, Edge AI Device, Security, Smart Retail, Surveillance - Translation Priorities: Optional Video Management Software (VMS) solutions are available in the market for more than a decade and some surveillance vendors are offering Network Video Recorders (NVR) devices preloaded with VMS. NVR with VMS controls multiple cameras and uses analytics for the detection and management of various events. The usage of advanced intelligent analytics in such systems is growing. Market researchers estimate CAGR (Compound Annual Growth Rate) of >20% for VMS in the coming 10 years. The key drivers for the growth include increasing security concerns and the adoption of VMS solutions in smart cities. Contributing to the significant growth is the rapid adoption of IP cameras for surveillance and security applications and the rising need for enhanced security. The VMS solutions are used for prevention of the personal and property crimes as well as for analytics that reveals valuable information on customers’ behavior. The efficiency of VMS solutions highly depends on the effectiveness of their intelligent analytics. Intelligent analytics are being used for real-time event detection and classification as well as for additional specific tasks like License Plate Recognition (LPR), behavioral detection, and face recognition. This blog post outlines the key challenges in meeting the requirements of small businesses and smart retail for a cost-effective surveillance solution with up to 6 IP surveillance cameras. While the number of cameras is limited in this segment, enhanced security with advanced intelligent analytics became a necessity and the customers expect the surveillance systems to detect more event types and achieve higher accuracy. In the... --- > KI-Videoanalysen sind eine kostengünstige Lösung, die eine Ereigniserkennung in Echtzeit für den intelligenten Einzelhandel ermöglicht. - Published: 2022-06-16 - Modified: 2024-04-25 - URL: https://hailo.ai/de/blog/advanced-video-analytics-for-small-businesses/ - Categories: Edge AI Box, Edge AI Device, Edge AI Device, Security, Security, Smart Retail, Surveillance, Surveillance - Translation Priorities: Optional Video Management Software (VMS) solutions are available in the market for more than a decade and some surveillance vendors are offering Network Video Recorders (NVR) devices preloaded with VMS. NVR with VMS controls multiple cameras and uses analytics for the detection and management of various events. The usage of advanced intelligent analytics in such systems is growing. Market researchers estimate CAGR (Compound Annual Growth Rate) of >20% for VMS in the coming 10 years. The key drivers for the growth include increasing security concerns and the adoption of VMS solutions in smart cities. Contributing to the significant growth is the rapid adoption of IP cameras for surveillance and security applications and the rising need for enhanced security. The VMS solutions are used for prevention of the personal and property crimes as well as for analytics that reveals valuable information on customers’ behavior. The efficiency of VMS solutions highly depends on the effectiveness of their intelligent analytics. Intelligent analytics are being used for real-time event detection and classification as well as for additional specific tasks like License Plate Recognition (LPR), behavioral detection, and face recognition. This blog post outlines the key challenges in meeting the requirements of small businesses and smart retail for a cost-effective surveillance solution with up to 6 IP surveillance cameras. While the number of cameras is limited in this segment, enhanced security with advanced intelligent analytics became a necessity and the customers expect the surveillance systems to detect more event types and achieve higher accuracy. In the... --- - Published: 2022-06-16 - Modified: 2024-11-05 - URL: https://hailo.ai/zh-hans/blog/advanced-video-analytics-for-small-businesses/ - Categories: Edge AI Box, Edge AI Device, Security, Smart Retail, Surveillance - Translation Priorities: 可选 Video Management Software (VMS) solutions are available in the market for more than a decade and some surveillance vendors are offering Network Video Recorders (NVR) devices preloaded with VMS. NVR with VMS controls multiple cameras and uses analytics for the detection and management of various events. The usage of advanced intelligent analytics in such systems is growing. Market researchers estimate CAGR (Compound Annual Growth Rate) of >20% for VMS in the coming 10 years. The key drivers for the growth include increasing security concerns and the adoption of VMS solutions in smart cities. Contributing to the significant growth is the rapid adoption of IP cameras for surveillance and security applications and the rising need for enhanced security. The VMS solutions are used for prevention of the personal and property crimes as well as for analytics that reveals valuable information on customers’ behavior. The efficiency of VMS solutions highly depends on the effectiveness of their intelligent analytics. Intelligent analytics are being used for real-time event detection and classification as well as for additional specific tasks like License Plate Recognition (LPR), behavioral detection, and face recognition. This blog post outlines the key challenges in meeting the requirements of small businesses and smart retail for a cost-effective surveillance solution with up to 6 IP surveillance cameras. While the number of cameras is limited in this segment, enhanced security with advanced intelligent analytics became a necessity and the customers expect the surveillance systems to detect more event types and achieve higher accuracy. In the... --- - Published: 2022-06-16 - Modified: 2024-11-05 - URL: https://hailo.ai/ja/blog/advanced-video-analytics-for-small-businesses/ - Categories: Edge AI Box, Edge AI Device, Security, Smart Retail, Surveillance - Translation Priorities: 可选 Video Management Software (VMS) solutions are available in the market for more than a decade and some surveillance vendors are offering Network Video Recorders (NVR) devices preloaded with VMS. NVR with VMS controls multiple cameras and uses analytics for the detection and management of various events. The usage of advanced intelligent analytics in such systems is growing. Market researchers estimate CAGR (Compound Annual Growth Rate) of >20% for VMS in the coming 10 years. The key drivers for the growth include increasing security concerns and the adoption of VMS solutions in smart cities. Contributing to the significant growth is the rapid adoption of IP cameras for surveillance and security applications and the rising need for enhanced security. The VMS solutions are used for prevention of the personal and property crimes as well as for analytics that reveals valuable information on customers’ behavior. The efficiency of VMS solutions highly depends on the effectiveness of their intelligent analytics. Intelligent analytics are being used for real-time event detection and classification as well as for additional specific tasks like License Plate Recognition (LPR), behavioral detection, and face recognition. This blog post outlines the key challenges in meeting the requirements of small businesses and smart retail for a cost-effective surveillance solution with up to 6 IP surveillance cameras. While the number of cameras is limited in this segment, enhanced security with advanced intelligent analytics became a necessity and the customers expect the surveillance systems to detect more event types and achieve higher accuracy. In the... --- > Hailo leads the way in traffic monitoring AI, transforming ITS into smarter, safer networks. Learn how in our detailed blog post. - Published: 2022-02-14 - Modified: 2024-11-05 - URL: https://hailo.ai/blog/powerful-edge-ai-to-boost-intelligent-traffic-systems-its/ - Categories: Smart City, Smart Transportation, Surveillance, Traffic Monitoring - Translation Priorities: Optional Whether in the city or outside of it, current traffic monitoring solutions have significant challenges in dealing with increasingly common complex traffic scenarios and more robust AI functionality requirements. Our cities are growing smarter as local and federal, private and public actors introduce AI and IoT technologies to better manage and plan them. Alongside energy efficiency, pollution reduction, and water and waste management, managing traffic and urban mobility are central pillars in what makes a Smart City “smart” (1). Traffic congestion plagues metropolitan areas across the world, impacting the environment and quality of life in these areas, as well as local and national economies. According to recent research by INRIX, time lost in traffic has cost the US economy $53 billion in 2021, while on the other side of the pond the UK and German economies lost £8 billion and €3. 5 billion, respectively (2). As with many complex problems, AI is increasingly applied to solve traffic, mobility, and transportation management issues (3). Traffic congestion is a complex problem that requires a multi-faceted approach where data collection and real-time monitoring are important tools. The former informs strategy and planning, while the latter provides the tactical knowledge that enables to take timely action on the ground. This is where edge-based Intelligent Video Analytics (IVA) comes in. What Smart Cameras and Intelligent Video Analytics Are Used For Urban areas are characterized by narrower spaces and a higher density of vehicles. While street speeds are relatively low (typically under 50 km/h), road occupancy is very high in specific bottleneck areas at specific times. Points of entry and exit of the city are such spaces, where major intercity arteries drain into narrower streets... --- - Published: 2022-02-14 - Modified: 2024-03-26 - URL: https://hailo.ai/de/blog/powerful-edge-ai-to-boost-intelligent-traffic-systems-its/ - Categories: Smart City, Smart City, Smart Transportation, Surveillance, Surveillance, Traffic Monitoring - Translation Priorities: Optional Whether in the city or outside of it, current traffic monitoring solutions have significant challenges in dealing with increasingly common complex traffic scenarios and more robust AI functionality requirements. Our cities are growing smarter as local and federal, private and public actors introduce AI and IoT technologies to better manage and plan them. Alongside energy efficiency, pollution reduction, and water and waste management, managing traffic and urban mobility are central pillars in what makes a Smart City “smart” (1). Traffic congestion plagues metropolitan areas across the world, impacting the environment and quality of life in these areas, as well as local and national economies. According to recent research by INRIX, time lost in traffic has cost the US economy $53 billion in 2021, while on the other side of the pond the UK and German economies lost £8 billion and €3. 5 billion, respectively (2). As with many complex problems, AI is increasingly applied to solve traffic, mobility, and transportation management issues (3). Traffic congestion is a complex problem that requires a multi-faceted approach where data collection and real-time monitoring are important tools. The former informs strategy and planning, while the latter provides the tactical knowledge that enables to take timely action on the ground. This is where edge-based Intelligent Video Analytics (IVA) comes in. What Smart Cameras and Intelligent Video Analytics Are Used For Urban areas are characterized by narrower spaces and a higher density of vehicles. While street speeds are relatively low (typically under 50 km/h), road occupancy is very high in specific bottleneck areas at specific times. Points of entry and exit of the city are such spaces, where major intercity arteries drain into narrower streets... --- - Published: 2022-02-14 - Modified: 2024-11-05 - URL: https://hailo.ai/zh-hans/blog/powerful-edge-ai-to-boost-intelligent-traffic-systems-its/ - Categories: Smart City, Smart Transportation, Surveillance, Traffic Monitoring - Translation Priorities: 可选 Whether in the city or outside of it, current traffic monitoring solutions have significant challenges in dealing with increasingly common complex traffic scenarios and more robust AI functionality requirements. Our cities are growing smarter as local and federal, private and public actors introduce AI and IoT technologies to better manage and plan them. Alongside energy efficiency, pollution reduction, and water and waste management, managing traffic and urban mobility are central pillars in what makes a Smart City “smart” (1). Traffic congestion plagues metropolitan areas across the world, impacting the environment and quality of life in these areas, as well as local and national economies. According to recent research by INRIX, time lost in traffic has cost the US economy $53 billion in 2021, while on the other side of the pond the UK and German economies lost £8 billion and €3. 5 billion, respectively (2). As with many complex problems, AI is increasingly applied to solve traffic, mobility, and transportation management issues (3). Traffic congestion is a complex problem that requires a multi-faceted approach where data collection and real-time monitoring are important tools. The former informs strategy and planning, while the latter provides the tactical knowledge that enables to take timely action on the ground. This is where edge-based Intelligent Video Analytics (IVA) comes in. What Smart Cameras and Intelligent Video Analytics Are Used For Urban areas are characterized by narrower spaces and a higher density of vehicles. While street speeds are relatively low (typically under 50 km/h), road occupancy is very high in specific bottleneck areas at specific times. Points of entry and exit of the city are such spaces, where major intercity arteries drain into narrower streets... --- - Published: 2022-02-14 - Modified: 2024-11-05 - URL: https://hailo.ai/ja/blog/powerful-edge-ai-to-boost-intelligent-traffic-systems-its/ - Categories: Smart City, Smart Transportation, Surveillance, Traffic Monitoring - Translation Priorities: 可选 Whether in the city or outside of it, current traffic monitoring solutions have significant challenges in dealing with increasingly common complex traffic scenarios and more robust AI functionality requirements. Our cities are growing smarter as local and federal, private and public actors introduce AI and IoT technologies to better manage and plan them. Alongside energy efficiency, pollution reduction, and water and waste management, managing traffic and urban mobility are central pillars in what makes a Smart City “smart” (1). Traffic congestion plagues metropolitan areas across the world, impacting the environment and quality of life in these areas, as well as local and national economies. According to recent research by INRIX, time lost in traffic has cost the US economy $53 billion in 2021, while on the other side of the pond the UK and German economies lost £8 billion and €3. 5 billion, respectively (2). As with many complex problems, AI is increasingly applied to solve traffic, mobility, and transportation management issues (3). Traffic congestion is a complex problem that requires a multi-faceted approach where data collection and real-time monitoring are important tools. The former informs strategy and planning, while the latter provides the tactical knowledge that enables to take timely action on the ground. This is where edge-based Intelligent Video Analytics (IVA) comes in. What Smart Cameras and Intelligent Video Analytics Are Used For Urban areas are characterized by narrower spaces and a higher density of vehicles. While street speeds are relatively low (typically under 50 km/h), road occupancy is very high in specific bottleneck areas at specific times. Points of entry and exit of the city are such spaces, where major intercity arteries drain into narrower streets... --- > Uncover the challenges of mass AI adoption, the role of vendor collaboration, and Hailo's innovative onboarding solutions. - Published: 2022-01-12 - Modified: 2025-05-05 - URL: https://hailo.ai/blog/how-software-can-streamline-customer-experience-in-edge-ai/ - Categories: AI Hardware, AI Software, Developer, Edge AI Developer, Edge AI Device - Translation Priorities: Optional When looking at the challenges facing mass AI adoption, infrastructure and integration are among their most important aspects. They were among the top limiting factors in a 2020 O’Rielly survey. When looking at the challenges facing mass AI adoption, infrastructure and integration are among their most important aspects. They were among the top limiting factors in a 2020 O’Rielly survey. The variety of frameworks, both for model generation and real-time deployment, also poses hurdles to ease and streamlining of development processes. In our experience, this is even more significant for successful edge AI adoption. As described by others, the customer will have to collaborate closely with its vendor to overcome integration challenges and make sure that everyone has a clear understanding of the process. This requires the vendor to have broader expertise beyond his core competencies. For edge AI use cases, a customer may need to combine a dedicated inference accelerator (such as that of the Hailo-8) with a host processor. He or she would also need to port their neural net model to a format the accelerator can work with, which, in some cases, may require some modifications to the model to reduce its size. This is just one example of the tight collaboration required to make an edge AI project work, especially at scale. Hailo has tackled and continues to tackle integration challenges on a regular basis. This has led us to pre-solve issues by developing several solutions and processes to simplify new customer onboarding and to support the problems existing customers may face when looking for an end-to-end edge AI solution. One of the first key factors to successful onboarding we have identified is the ease of... --- > Entdecken Sie die Herausforderungen der Masseneinführung von KI, die Rolle der Zusammenarbeit mit Anbietern und die innovativen von Hailo. - Published: 2022-01-12 - Modified: 2024-10-07 - URL: https://hailo.ai/de/blog/how-software-can-streamline-customer-experience-in-edge-ai/ - Categories: AI Hardware, AI Hardware, AI Software, AI Software, Developer, Edge AI Developer, Edge AI Developer, Edge AI Device, Edge AI Device - Translation Priorities: Optional When looking at the challenges facing mass AI adoption, infrastructure and integration are among their most important aspects. They were among the top limiting factors in a 2020 O’Rielly survey. When looking at the challenges facing mass AI adoption, infrastructure and integration are among their most important aspects. They were among the top limiting factors in a 2020 O’Rielly survey. The variety of frameworks, both for model generation and real-time deployment, also poses hurdles to ease and streamlining of development processes. In our experience, this is even more significant for successful edge AI adoption. As described by others, the customer will have to collaborate closely with its vendor to overcome integration challenges and make sure that everyone has a clear understanding of the process. This requires the vendor to have broader expertise beyond his core competencies. For edge AI use cases, a customer may need to combine a dedicated inference accelerator (such as that of the Hailo-8) with a host processor. He or she would also need to port their neural net model to a format the accelerator can work with, which, in some cases, may require some modifications to the model to reduce its size. This is just one example of the tight collaboration required to make an edge AI project work, especially at scale. Hailo has tackled and continues to tackle integration challenges on a regular basis. This has led us to pre-solve issues by developing several solutions and processes to simplify new customer onboarding and to support the problems existing customers may face when looking for an end-to-end edge AI solution. One of the first key factors to successful onboarding we have identified is the ease of... --- - Published: 2022-01-12 - Modified: 2024-11-06 - URL: https://hailo.ai/zh-hans/blog/how-software-can-streamline-customer-experience-in-edge-ai/ - Categories: AI Hardware, AI Software, Developer, Edge AI Developer, Edge AI Device - Translation Priorities: 可选 When looking at the challenges facing mass AI adoption, infrastructure and integration are among their most important aspects. They were among the top limiting factors in a 2020 O’Rielly survey. When looking at the challenges facing mass AI adoption, infrastructure and integration are among their most important aspects. They were among the top limiting factors in a 2020 O’Rielly survey. The variety of frameworks, both for model generation and real-time deployment, also poses hurdles to ease and streamlining of development processes. In our experience, this is even more significant for successful edge AI adoption. As described by others, the customer will have to collaborate closely with its vendor to overcome integration challenges and make sure that everyone has a clear understanding of the process. This requires the vendor to have broader expertise beyond his core competencies. For edge AI use cases, a customer may need to combine a dedicated inference accelerator (such as that of the Hailo-8) with a host processor. He or she would also need to port their neural net model to a format the accelerator can work with, which, in some cases, may require some modifications to the model to reduce its size. This is just one example of the tight collaboration required to make an edge AI project work, especially at scale. Hailo has tackled and continues to tackle integration challenges on a regular basis. This has led us to pre-solve issues by developing several solutions and processes to simplify new customer onboarding and to support the problems existing customers may face when looking for an end-to-end edge AI solution. One of the first key factors to successful onboarding we have identified is the ease of... --- - Published: 2022-01-12 - Modified: 2025-05-05 - URL: https://hailo.ai/ja/blog/how-software-can-streamline-customer-experience-in-edge-ai/ - Categories: AI Hardware, AI Software, Developer, Edge AI Developer, Edge AI Device - Translation Priorities: 可选 When looking at the challenges facing mass AI adoption, infrastructure and integration are among their most important aspects. They were among the top limiting factors in a 2020 O’Rielly survey. When looking at the challenges facing mass AI adoption, infrastructure and integration are among their most important aspects. They were among the top limiting factors in a 2020 O’Rielly survey. The variety of frameworks, both for model generation and real-time deployment, also poses hurdles to ease and streamlining of development processes. In our experience, this is even more significant for successful edge AI adoption. As described by others, the customer will have to collaborate closely with its vendor to overcome integration challenges and make sure that everyone has a clear understanding of the process. This requires the vendor to have broader expertise beyond his core competencies. For edge AI use cases, a customer may need to combine a dedicated inference accelerator (such as that of the Hailo-8) with a host processor. He or she would also need to port their neural net model to a format the accelerator can work with, which, in some cases, may require some modifications to the model to reduce its size. This is just one example of the tight collaboration required to make an edge AI project work, especially at scale. Hailo has tackled and continues to tackle integration challenges on a regular basis. This has led us to pre-solve issues by developing several solutions and processes to simplify new customer onboarding and to support the problems existing customers may face when looking for an end-to-end edge AI solution. One of the first key factors to successful onboarding we have identified is the ease of... --- > Join Hailo's CBO in exploring the edge AI landscape, its challenges, and the company's role in shaping its future. - Published: 2021-12-22 - Modified: 2024-11-05 - URL: https://hailo.ai/blog/5-questions-on-the-edge-ai-with-hailo-cbo-hadar-zeitlin/ - Categories: AI Hardware, AI Software, Automotive, Edge AI Device - Translation Priorities: Optional It is very exciting to see the increasing adoption of AI in edge applications in recent years and how the ecosystem is being built to further benefit from AI capabilities by implementing more sophisticated, high performance use cases. Where are you seeing the most demand for high-performance edge AI? It is very exciting to see the increasing adoption of AI in edge applications in recent years and how the ecosystem is being built to further benefit from AI capabilities by implementing more sophisticated, high performance use cases. I see this happening across-the-board. ADAS/AV applications are one prominent example. These rely more and more on multiple sensors and a significant portion of them use high resolution cameras, which are data intense. The applications are sensitive to latency (short reaction time) and quality (low false alarm and misdetection rates), which means that high accuracy and high FPS requirements are driving the need for high compute. In the Security & Public Safety domain, I see high-performance AI especially necessary both in cameras and video processing devices. In cameras, customers are looking to process high-resolution video using high-accuracy algorithms. This translates into wider effective coverage in the camera (which means they can deploy fewer cameras to cover the same RoI), as well as reduction in false alerts (a better product producing better results). Customers are also looking to run multiple applications on the same stream (for example monitoring occupancy in parallel to activity detection), which also drives the demand for greater compute. In addition to cameras, there are systems processing multiple video streams. These are, typically, either an added smart aggregation point in an existing “non-AI” camera deployment or new deployments with centralized architecture at the edge with the goal of reducing... --- > Entdecken Sie gemeinsam mit dem CBO von Hailo die Edge-KI und die Rolle des Unternehmens bei der Gestaltung seiner Zukunft. - Published: 2021-12-22 - Modified: 2024-04-25 - URL: https://hailo.ai/de/blog/5-questions-on-the-edge-ai-with-hailo-cbo-hadar-zeitlin/ - Categories: AI Hardware, AI Hardware, AI Software, Automotive, Automotive, Edge AI Device, Edge AI Device - Translation Priorities: Optional It is very exciting to see the increasing adoption of AI in edge applications in recent years and how the ecosystem is being built to further benefit from AI capabilities by implementing more sophisticated, high performance use cases. Where are you seeing the most demand for high-performance edge AI? It is very exciting to see the increasing adoption of AI in edge applications in recent years and how the ecosystem is being built to further benefit from AI capabilities by implementing more sophisticated, high performance use cases. I see this happening across-the-board. ADAS/AV applications are one prominent example. These rely more and more on multiple sensors and a significant portion of them use high resolution cameras, which are data intense. The applications are sensitive to latency (short reaction time) and quality (low false alarm and misdetection rates), which means that high accuracy and high FPS requirements are driving the need for high compute. In the Security & Public Safety domain, I see high-performance AI especially necessary both in cameras and video processing devices. In cameras, customers are looking to process high-resolution video using high-accuracy algorithms. This translates into wider effective coverage in the camera (which means they can deploy fewer cameras to cover the same RoI), as well as reduction in false alerts (a better product producing better results). Customers are also looking to run multiple applications on the same stream (for example monitoring occupancy in parallel to activity detection), which also drives the demand for greater compute. In addition to cameras, there are systems processing multiple video streams. These are, typically, either an added smart aggregation point in an existing “non-AI” camera deployment or new deployments with centralized architecture at the edge with the goal of reducing... --- - Published: 2021-12-22 - Modified: 2024-11-05 - URL: https://hailo.ai/zh-hans/blog/5-questions-on-the-edge-ai-with-hailo-cbo-hadar-zeitlin/ - Categories: AI Hardware, AI Software, Automotive, Edge AI Device - Translation Priorities: 可选 It is very exciting to see the increasing adoption of AI in edge applications in recent years and how the ecosystem is being built to further benefit from AI capabilities by implementing more sophisticated, high performance use cases. Where are you seeing the most demand for high-performance edge AI? It is very exciting to see the increasing adoption of AI in edge applications in recent years and how the ecosystem is being built to further benefit from AI capabilities by implementing more sophisticated, high performance use cases. I see this happening across-the-board. ADAS/AV applications are one prominent example. These rely more and more on multiple sensors and a significant portion of them use high resolution cameras, which are data intense. The applications are sensitive to latency (short reaction time) and quality (low false alarm and misdetection rates), which means that high accuracy and high FPS requirements are driving the need for high compute. In the Security & Public Safety domain, I see high-performance AI especially necessary both in cameras and video processing devices. In cameras, customers are looking to process high-resolution video using high-accuracy algorithms. This translates into wider effective coverage in the camera (which means they can deploy fewer cameras to cover the same RoI), as well as reduction in false alerts (a better product producing better results). Customers are also looking to run multiple applications on the same stream (for example monitoring occupancy in parallel to activity detection), which also drives the demand for greater compute. In addition to cameras, there are systems processing multiple video streams. These are, typically, either an added smart aggregation point in an existing “non-AI” camera deployment or new deployments with centralized architecture at the edge with the goal of reducing... --- > Join Hailo's CBO in exploring the edge AI landscape, its challenges, and the company's role in shaping its future. - Published: 2021-12-22 - Modified: 2024-11-05 - URL: https://hailo.ai/ja/blog/5-questions-on-the-edge-ai-with-hailo-cbo-hadar-zeitlin/ - Categories: AI Hardware, AI Software, Automotive, Edge AI Device - Translation Priorities: 可选 It is very exciting to see the increasing adoption of AI in edge applications in recent years and how the ecosystem is being built to further benefit from AI capabilities by implementing more sophisticated, high performance use cases. Where are you seeing the most demand for high-performance edge AI? It is very exciting to see the increasing adoption of AI in edge applications in recent years and how the ecosystem is being built to further benefit from AI capabilities by implementing more sophisticated, high performance use cases. I see this happening across-the-board. ADAS/AV applications are one prominent example. These rely more and more on multiple sensors and a significant portion of them use high resolution cameras, which are data intense. The applications are sensitive to latency (short reaction time) and quality (low false alarm and misdetection rates), which means that high accuracy and high FPS requirements are driving the need for high compute. In the Security & Public Safety domain, I see high-performance AI especially necessary both in cameras and video processing devices. In cameras, customers are looking to process high-resolution video using high-accuracy algorithms. This translates into wider effective coverage in the camera (which means they can deploy fewer cameras to cover the same RoI), as well as reduction in false alerts (a better product producing better results). Customers are also looking to run multiple applications on the same stream (for example monitoring occupancy in parallel to activity detection), which also drives the demand for greater compute. In addition to cameras, there are systems processing multiple video streams. These are, typically, either an added smart aggregation point in an existing “non-AI” camera deployment or new deployments with centralized architecture at the edge with the goal of reducing... --- > Intelligent Video Analytics: Learn how edge AI is becoming mainstream, providing real-time video analytics on intelligent cameras, NVRs, and edge AI boxes. - Published: 2021-12-09 - Modified: 2024-11-05 - URL: https://hailo.ai/blog/a-new-generation-of-video-analytics-enabled-by-powerful-edge-ai/ - Categories: Edge AI Device, Intelligent Camera, Security, Smart Retail, Surveillance - Translation Priorities: Optional Intelligent Video Analytics (IVA) have penetrated many applications across industry verticals. Starting from on-premises server-based capabilities, then becoming available on the cloud (relying on high-bandwidth connectivity) and, more recently, on edge devices. Intelligent Video Analytics (IVA) have penetrated many applications across industry verticals. Starting from on-premises server-based capabilities, then becoming available on the cloud (relying on high-bandwidth connectivity) and, more recently, on edge devices. In this respect, “edge” includes intelligent cameras, intelligent NVRs (Network Video Recorders) and small dedicated appliances (often referred to as edge AI boxes or video analytics boxes) with built-in AI processing. Today, AI-based on-device analytics solutions abound and their number is growing fast. Research leader Omdia estimated that in 2021, 26% percent of cameras and NVRs sold had AI capabilities. They forecast that by 2025 this share will reach over 60%, with AI-capable cameras making up 64% of all IP camera shipped worldwide. Edge AI is now entering the mainstream market and becoming widely available in mid-range solutions. Edge AI has made it possible for cameras and other small devices to recognize objects and people, track movement and even identify behavior. However, these capabilities first arrived at the edge in a limited capacity and, as it happens with many budding technologies, disappointed users with lackluster performance: low-quality detections and tracking and unreliable alerts. Analytics quality, which has gradually improved since those early days, is currently taking a big leap forward. A new generation of AI SoCs is starting to see adoption and bringing more powerful, data center-level processing capabilities to the edge. Why do you need more AI TOPS (tera operations per second) on your camera or device? What is the point of being able to run a... --- > Intelligente Videoanalyse: Erfahren Sie, wie Edge AI Echtzeit-Videoanalysen für intelligente Kameras, NVRs und Edge AI-Boxen bereitstellt. - Published: 2021-12-09 - Modified: 2024-04-25 - URL: https://hailo.ai/de/blog/a-new-generation-of-video-analytics-enabled-by-powerful-edge-ai/ - Categories: Edge AI Device, Edge AI Device, Intelligent Camera, Intelligent Camera, Security, Security, Smart Retail, Surveillance, Surveillance - Translation Priorities: Optional Intelligent Video Analytics (IVA) have penetrated many applications across industry verticals. Starting from on-premises server-based capabilities, then becoming available on the cloud (relying on high-bandwidth connectivity) and, more recently, on edge devices. Intelligent Video Analytics (IVA) have penetrated many applications across industry verticals. Starting from on-premises server-based capabilities, then becoming available on the cloud (relying on high-bandwidth connectivity) and, more recently, on edge devices. In this respect, “edge” includes intelligent cameras, intelligent NVRs (Network Video Recorders) and small dedicated appliances (often referred to as edge AI boxes or video analytics boxes) with built-in AI processing. Today, AI-based on-device analytics solutions abound and their number is growing fast. Research leader Omdia estimated that in 2021, 26% percent of cameras and NVRs sold had AI capabilities. They forecast that by 2025 this share will reach over 60%, with AI-capable cameras making up 64% of all IP camera shipped worldwide. Edge AI is now entering the mainstream market and becoming widely available in mid-range solutions. Edge AI has made it possible for cameras and other small devices to recognize objects and people, track movement and even identify behavior. However, these capabilities first arrived at the edge in a limited capacity and, as it happens with many budding technologies, disappointed users with lackluster performance: low-quality detections and tracking and unreliable alerts. Analytics quality, which has gradually improved since those early days, is currently taking a big leap forward. A new generation of AI SoCs is starting to see adoption and bringing more powerful, data center-level processing capabilities to the edge. Why do you need more AI TOPS (tera operations per second) on your camera or device? What is the point of being able to run a... --- - Published: 2021-12-09 - Modified: 2024-11-05 - URL: https://hailo.ai/zh-hans/blog/a-new-generation-of-video-analytics-enabled-by-powerful-edge-ai/ - Categories: Edge AI Device, Intelligent Camera, Security, Smart Retail, Surveillance - Translation Priorities: 可选 Intelligent Video Analytics (IVA) have penetrated many applications across industry verticals. Starting from on-premises server-based capabilities, then becoming available on the cloud (relying on high-bandwidth connectivity) and, more recently, on edge devices. Intelligent Video Analytics (IVA) have penetrated many applications across industry verticals. Starting from on-premises server-based capabilities, then becoming available on the cloud (relying on high-bandwidth connectivity) and, more recently, on edge devices. In this respect, “edge” includes intelligent cameras, intelligent NVRs (Network Video Recorders) and small dedicated appliances (often referred to as edge AI boxes or video analytics boxes) with built-in AI processing. Today, AI-based on-device analytics solutions abound and their number is growing fast. Research leader Omdia estimated that in 2021, 26% percent of cameras and NVRs sold had AI capabilities. They forecast that by 2025 this share will reach over 60%, with AI-capable cameras making up 64% of all IP camera shipped worldwide. Edge AI is now entering the mainstream market and becoming widely available in mid-range solutions. Edge AI has made it possible for cameras and other small devices to recognize objects and people, track movement and even identify behavior. However, these capabilities first arrived at the edge in a limited capacity and, as it happens with many budding technologies, disappointed users with lackluster performance: low-quality detections and tracking and unreliable alerts. Analytics quality, which has gradually improved since those early days, is currently taking a big leap forward. A new generation of AI SoCs is starting to see adoption and bringing more powerful, data center-level processing capabilities to the edge. Why do you need more AI TOPS (tera operations per second) on your camera or device? What is the point of being able to run a... --- - Published: 2021-12-09 - Modified: 2024-11-05 - URL: https://hailo.ai/ja/blog/a-new-generation-of-video-analytics-enabled-by-powerful-edge-ai/ - Categories: Edge AI Device, Intelligent Camera, Security, Smart Retail, Surveillance - Translation Priorities: 可选 Intelligent Video Analytics (IVA) have penetrated many applications across industry verticals. Starting from on-premises server-based capabilities, then becoming available on the cloud (relying on high-bandwidth connectivity) and, more recently, on edge devices. Intelligent Video Analytics (IVA) have penetrated many applications across industry verticals. Starting from on-premises server-based capabilities, then becoming available on the cloud (relying on high-bandwidth connectivity) and, more recently, on edge devices. In this respect, “edge” includes intelligent cameras, intelligent NVRs (Network Video Recorders) and small dedicated appliances (often referred to as edge AI boxes or video analytics boxes) with built-in AI processing. Today, AI-based on-device analytics solutions abound and their number is growing fast. Research leader Omdia estimated that in 2021, 26% percent of cameras and NVRs sold had AI capabilities. They forecast that by 2025 this share will reach over 60%, with AI-capable cameras making up 64% of all IP camera shipped worldwide. Edge AI is now entering the mainstream market and becoming widely available in mid-range solutions. Edge AI has made it possible for cameras and other small devices to recognize objects and people, track movement and even identify behavior. However, these capabilities first arrived at the edge in a limited capacity and, as it happens with many budding technologies, disappointed users with lackluster performance: low-quality detections and tracking and unreliable alerts. Analytics quality, which has gradually improved since those early days, is currently taking a big leap forward. A new generation of AI SoCs is starting to see adoption and bringing more powerful, data center-level processing capabilities to the edge. Why do you need more AI TOPS (tera operations per second) on your camera or device? What is the point of being able to run a... --- > Learn about Hailo and NXP's partnership in developing efficient embedded AI solutions. Explore our innovative technologies. Read more now! - Published: 2021-11-18 - Modified: 2024-11-05 - URL: https://hailo.ai/blog/play-to-win-how-hailo-and-nxp-create-efficient-embedded-ai-solutions/ - Categories: Automotive, Edge AI Developer, Edge AI Device, Industry 4.0, Intelligent Camera, Smart Transportation, Surveillance - Translation Priorities: Optional A good partnership is all about the added value. It highlights the strengths of each party, compensates for the weaker points and overall produces more value than what the two parties would produce individually. When I was a kid playing out in the field or somewhere around the neighborhood, I remember that choosing your teammates made all the difference - whether you won or lost and if you had any fun playing the game.  I also remember I had that one friend... We were inseparable and, even though every new game would start the team selection process anew, we were always on the same team.   Choosing the right teammates is just as important in grownup life and, like then, it is what makes the difference between winning and losing.  A good partnership is all about the added value. It highlights the strengths of each party, compensates for the weaker points, and overall produces more value than what the two parties would produce individually.   Our recent partnership with NXP is a good example of 1 + 1 > 2 and where good partners make all the difference.  Integrating the powerful Hailo-8 AI processor into a range of NXP platforms has created some amazing computing solutions across applications and industries.  To streamline the use of these mature, power-efficient, and cost-effective solutions, we have all the runtime software and a set of full-fledged GStreamer-based apps ready for deployment.   Let’s look at some example use cases made possible or empowered by this partnership.  Some of these are actual applications being developed by our partners, while others are cool applications unlocked by powerful AI processing at the edge.   Smart Transportation: Train Occupancy  Standing on the platform when a train rolls into the station, there is always a hint of a gamble when you try to choose the cart to enter. You try to peek through the doors and windows to gauge which cart has fewer passengers and maybe even an open seat or two. Wouldn’t it be nicer if the train could just “tell” you which carts are less crowded or have vacant seats ahead of time?   The combination of the NXP i. MX... --- > Visit Hailo's blog to learn how integrating the powerful Hailo-8 AI processor into a range of NXP platforms has created some amazing computing solutions across applications and industries. - Published: 2021-11-18 - Modified: 2024-06-27 - URL: https://hailo.ai/de/blog/play-to-win-how-hailo-and-nxp-create-efficient-embedded-ai-solutions/ - Categories: Automotive, Automotive, Edge AI Developer, Edge AI Developer, Edge AI Device, Edge AI Device, Industry 4.0, Intelligent Camera, Intelligent Camera, Smart Transportation, Surveillance, Surveillance - Translation Priorities: Optional A good partnership is all about the added value. It highlights the strengths of each party, compensates for the weaker points and overall produces more value than what the two parties would produce individually. When I was a kid playing out in the field or somewhere around the neighborhood, I remember that choosing your teammates made all the difference - whether you won or lost and if you had any fun playing the game.  I also remember I had that one friend... We were inseparable and, even though every new game would start the team selection process anew, we were always on the same team.   Choosing the right teammates is just as important in grownup life and, like then, it is what makes the difference between winning and losing.  A good partnership is all about the added value. It highlights the strengths of each party, compensates for the weaker points, and overall produces more value than what the two parties would produce individually.   Our recent partnership with NXP is a good example of 1 + 1 > 2 and where good partners make all the difference.  Integrating the powerful Hailo-8 AI processor into a range of NXP platforms has created some amazing computing solutions across applications and industries.  To streamline the use of these mature, power-efficient, and cost-effective solutions, we have all the runtime software and a set of full-fledged GStreamer-based apps ready for deployment.   Let’s look at some example use cases made possible or empowered by this partnership.  Some of these are actual applications being developed by our partners, while others are cool applications unlocked by powerful AI processing at the edge.   Smart Transportation: Train Occupancy  Standing on the platform when a train rolls into the station, there is always a hint of a gamble when you try to choose the cart to enter. You try to peek through the doors and windows to gauge which cart has fewer passengers and maybe even an open seat or two. Wouldn’t it be nicer if the train could just “tell” you which carts are less crowded or have vacant seats ahead of time?   The combination of the NXP i. MX... --- - Published: 2021-11-18 - Modified: 2024-11-05 - URL: https://hailo.ai/zh-hans/blog/play-to-win-how-hailo-and-nxp-create-efficient-embedded-ai-solutions/ - Categories: Automotive, Edge AI Developer, Edge AI Device, Industry 4.0, Intelligent Camera, Smart Transportation, Surveillance - Translation Priorities: 可选 A good partnership is all about the added value. It highlights the strengths of each party, compensates for the weaker points and overall produces more value than what the two parties would produce individually. When I was a kid playing out in the field or somewhere around the neighborhood, I remember that choosing your teammates made all the difference - whether you won or lost and if you had any fun playing the game.  I also remember I had that one friend... We were inseparable and, even though every new game would start the team selection process anew, we were always on the same team.   Choosing the right teammates is just as important in grownup life and, like then, it is what makes the difference between winning and losing.  A good partnership is all about the added value. It highlights the strengths of each party, compensates for the weaker points, and overall produces more value than what the two parties would produce individually.   Our recent partnership with NXP is a good example of 1 + 1 > 2 and where good partners make all the difference.  Integrating the powerful Hailo-8 AI processor into a range of NXP platforms has created some amazing computing solutions across applications and industries.  To streamline the use of these mature, power-efficient, and cost-effective solutions, we have all the runtime software and a set of full-fledged GStreamer-based apps ready for deployment.   Let’s look at some example use cases made possible or empowered by this partnership.  Some of these are actual applications being developed by our partners, while others are cool applications unlocked by powerful AI processing at the edge.   Smart Transportation: Train Occupancy  Standing on the platform when a train rolls into the station, there is always a hint of a gamble when you try to choose the cart to enter. You try to peek through the doors and windows to gauge which cart has fewer passengers and maybe even an open seat or two. Wouldn’t it be nicer if the train could just “tell” you which carts are less crowded or have vacant seats ahead of time?   The combination of the NXP i. MX... --- - Published: 2021-11-18 - Modified: 2024-11-05 - URL: https://hailo.ai/ja/blog/play-to-win-how-hailo-and-nxp-create-efficient-embedded-ai-solutions/ - Categories: Automotive, Edge AI Developer, Edge AI Device, Industry 4.0, Intelligent Camera, Smart Transportation, Surveillance - Translation Priorities: 可选 A good partnership is all about the added value. It highlights the strengths of each party, compensates for the weaker points and overall produces more value than what the two parties would produce individually. When I was a kid playing out in the field or somewhere around the neighborhood, I remember that choosing your teammates made all the difference - whether you won or lost and if you had any fun playing the game.  I also remember I had that one friend... We were inseparable and, even though every new game would start the team selection process anew, we were always on the same team.   Choosing the right teammates is just as important in grownup life and, like then, it is what makes the difference between winning and losing.  A good partnership is all about the added value. It highlights the strengths of each party, compensates for the weaker points, and overall produces more value than what the two parties would produce individually.   Our recent partnership with NXP is a good example of 1 + 1 > 2 and where good partners make all the difference.  Integrating the powerful Hailo-8 AI processor into a range of NXP platforms has created some amazing computing solutions across applications and industries.  To streamline the use of these mature, power-efficient, and cost-effective solutions, we have all the runtime software and a set of full-fledged GStreamer-based apps ready for deployment.   Let’s look at some example use cases made possible or empowered by this partnership.  Some of these are actual applications being developed by our partners, while others are cool applications unlocked by powerful AI processing at the edge.   Smart Transportation: Train Occupancy  Standing on the platform when a train rolls into the station, there is always a hint of a gamble when you try to choose the cart to enter. You try to peek through the doors and windows to gauge which cart has fewer passengers and maybe even an open seat or two. Wouldn’t it be nicer if the train could just “tell” you which carts are less crowded or have vacant seats ahead of time?   The combination of the NXP i. MX... --- > Hailo is revolutionizing ADAS sensing by integrating AI for enhanced efficiency. From sensor capabilities to real-world applications. - Published: 2021-10-27 - Modified: 2024-11-05 - URL: https://hailo.ai/blog/pairing-sensing-with-ai-for-efficient-adas/ - Categories: ADAS, AI Benchmarks, AI Hardware, Automotive, Object Detection - Translation Priorities: Optional Sensor and sensing capabilities are common and expanding in modern vehicles. One of the major motivators for this is safety or, more specifically, Euro NCAP Vision Zero guidelines. These are driving automakers’ and Tier 1 suppliers’ ADAS/AV roadmaps, requiring more powerful scene understanding. Among other things, extending the scope of VRU (vulnerable road users) protection is driving up sensor variety, resolution and quality and, as a result, poses increasingly high vision processing requirements. Addressing this need for higher processing throughput (within the power and space limitations of the vehicle) is now made possible by novel highly power efficient AI processors. The Forward-Facing Camera At the basic SAE Level 2, let’s look at the single-sensor forward-facing-camera (FFC). The capability of an intelligent FFCs to offer VRU protection relies on its sensor and the AI processing capabilities. It is a vehicle safety requirement to have a high-resolution sensor to be able to capture even smaller objects like pedestrians, animals and road signs, while in motion. The high resolution is also a prerequisite for safety at higher driving speeds, which dictate an extended braking distance. High resolutions demand more from the camera’s AI processing. To put this in more practical terms: a car travelling at 80 km/h (or 50 MPH) needs to be able to detect a child crossing the road. Let’s assume the child’s horizontal profile is 60 cm (24 inches). A typical object detection neural network will require at least 8 pixels for proper detection and identification of the child. Also,... --- > Hailo revolutioniert die ADAS-Sensorik durch die Integration von KI für mehr Effizienz. Von Sensorfunktionen bis hin zu realen Anwendungen. - Published: 2021-10-27 - Modified: 2024-06-13 - URL: https://hailo.ai/de/blog/pairing-sensing-with-ai-for-efficient-adas/ - Categories: ADAS, ADAS, AI Benchmarks, AI Hardware, AI Hardware, Automotive, Automotive, Object Detection - Translation Priorities: Optional Sensor and sensing capabilities are common and expanding in modern vehicles. One of the major motivators for this is safety or, more specifically, Euro NCAP Vision Zero guidelines. These are driving automakers’ and Tier 1 suppliers’ ADAS/AV roadmaps, requiring more powerful scene understanding. Among other things, extending the scope of VRU (vulnerable road users) protection is driving up sensor variety, resolution and quality and, as a result, poses increasingly high vision processing requirements. Addressing this need for higher processing throughput (within the power and space limitations of the vehicle) is now made possible by novel highly power efficient AI processors. The Forward-Facing Camera At the basic SAE Level 2, let’s look at the single-sensor forward-facing-camera (FFC). The capability of an intelligent FFCs to offer VRU protection relies on its sensor and the AI processing capabilities. It is a vehicle safety requirement to have a high-resolution sensor to be able to capture even smaller objects like pedestrians, animals and road signs, while in motion. The high resolution is also a prerequisite for safety at higher driving speeds, which dictate an extended braking distance. High resolutions demand more from the camera’s AI processing. To put this in more practical terms: a car travelling at 80 km/h (or 50 MPH) needs to be able to detect a child crossing the road. Let’s assume the child’s horizontal profile is 60 cm (24 inches). A typical object detection neural network will require at least 8 pixels for proper detection and identification of the child. Also,... --- - Published: 2021-10-27 - Modified: 2024-11-05 - URL: https://hailo.ai/zh-hans/blog/pairing-sensing-with-ai-for-efficient-adas/ - Categories: ADAS, AI Benchmarks, AI Hardware, Automotive, Object Detection - Translation Priorities: 可选 Sensor and sensing capabilities are common and expanding in modern vehicles. One of the major motivators for this is safety or, more specifically, Euro NCAP Vision Zero guidelines. These are driving automakers’ and Tier 1 suppliers’ ADAS/AV roadmaps, requiring more powerful scene understanding. Among other things, extending the scope of VRU (vulnerable road users) protection is driving up sensor variety, resolution and quality and, as a result, poses increasingly high vision processing requirements. Addressing this need for higher processing throughput (within the power and space limitations of the vehicle) is now made possible by novel highly power efficient AI processors. The Forward-Facing Camera At the basic SAE Level 2, let’s look at the single-sensor forward-facing-camera (FFC). The capability of an intelligent FFCs to offer VRU protection relies on its sensor and the AI processing capabilities. It is a vehicle safety requirement to have a high-resolution sensor to be able to capture even smaller objects like pedestrians, animals and road signs, while in motion. The high resolution is also a prerequisite for safety at higher driving speeds, which dictate an extended braking distance. High resolutions demand more from the camera’s AI processing. To put this in more practical terms: a car travelling at 80 km/h (or 50 MPH) needs to be able to detect a child crossing the road. Let’s assume the child’s horizontal profile is 60 cm (24 inches). A typical object detection neural network will require at least 8 pixels for proper detection and identification of the child. Also,... --- > Hailo is revolutionizing ADAS sensing by integrating AI for enhanced efficiency. From sensor capabilities to real-world applications. - Published: 2021-10-27 - Modified: 2024-11-05 - URL: https://hailo.ai/ja/blog/pairing-sensing-with-ai-for-efficient-adas/ - Categories: ADAS, AI Benchmarks, AI Hardware, Automotive, Object Detection - Translation Priorities: 可选 Sensor and sensing capabilities are common and expanding in modern vehicles. One of the major motivators for this is safety or, more specifically, Euro NCAP Vision Zero guidelines. These are driving automakers’ and Tier 1 suppliers’ ADAS/AV roadmaps, requiring more powerful scene understanding. Among other things, extending the scope of VRU (vulnerable road users) protection is driving up sensor variety, resolution and quality and, as a result, poses increasingly high vision processing requirements. Addressing this need for higher processing throughput (within the power and space limitations of the vehicle) is now made possible by novel highly power efficient AI processors. The Forward-Facing Camera At the basic SAE Level 2, let’s look at the single-sensor forward-facing-camera (FFC). The capability of an intelligent FFCs to offer VRU protection relies on its sensor and the AI processing capabilities. It is a vehicle safety requirement to have a high-resolution sensor to be able to capture even smaller objects like pedestrians, animals and road signs, while in motion. The high resolution is also a prerequisite for safety at higher driving speeds, which dictate an extended braking distance. High resolutions demand more from the camera’s AI processing. To put this in more practical terms: a car travelling at 80 km/h (or 50 MPH) needs to be able to detect a child crossing the road. Let’s assume the child’s horizontal profile is 60 cm (24 inches). A typical object detection neural network will require at least 8 pixels for proper detection and identification of the child. Also,... --- > Join Hailo CEO Orr Danon as he addresses five pivotal questions on Edge AI's impact, with insights on automotive autonomy, security, & more. - Published: 2021-09-22 - Modified: 2024-11-05 - URL: https://hailo.ai/blog/5-questions-on-the-edge-ai-with-hailo-ceo-orr-danon/ - Categories: AI Software, Edge AI Device, Intelligent Camera, Power Efficiency, Surveillance - Translation Priorities: Optional Where do you see high-performance Edge AI making the most impact? Automotive autonomy and driver assistance (ADAS) immediately comes to mind. To make the roads safer we increasingly let technology take the wheel – whether it’s supporting drivers or replacing them altogether. This takes a lot of AI (Artificial Intelligence), which translates into a lot of AI processing power that needs to be very efficient and be spread throughout the entire vehicle, including the vehicle edge. We saw a natural fit for our architecture here, so we designed an automotive-grade chip. Euro NCAP Vision Zero is something we feel strongly about, and our technology helps implement this vision in everyday cars. Security is another field where AI is creating enormous value. It might not be a revolutionary leap like in Automotive, but AI is penetrating in a big way. AI for surveillance and security operations – private, commercial and governmental – are all looking for better quality and efficiency gains. You don’t want to miss a single intruder, tailgater or emergency situation, and you certainly don’t want to spend your whole day going through false alarm notifications. You want to be able to react and even prevent bad things from happening. That is where AI comes in and people’s appetite for it is just getting bigger. Nowadays, they don’t just want the camera and the neural network to be there, they have an expectation of how they should perform: real-time, high-accuracy and for a wide coverage area. They require significant... --- > Begleiten Sie Orr Danon, CEO von Hailo, während er fünf zentrale Fragen zum Einfluss von Edge AI beantwortet. - Published: 2021-09-22 - Modified: 2024-06-23 - URL: https://hailo.ai/de/blog/5-questions-on-the-edge-ai-with-hailo-ceo-orr-danon/ - Categories: AI Software, Edge AI Device, Edge AI Device, Intelligent Camera, Intelligent Camera, Power Efficiency, Surveillance, Surveillance - Translation Priorities: Optional Where do you see high-performance Edge AI making the most impact? Automotive autonomy and driver assistance (ADAS) immediately comes to mind. To make the roads safer we increasingly let technology take the wheel – whether it’s supporting drivers or replacing them altogether. This takes a lot of AI (Artificial Intelligence), which translates into a lot of AI processing power that needs to be very efficient and be spread throughout the entire vehicle, including the vehicle edge. We saw a natural fit for our architecture here, so we designed an automotive-grade chip. Euro NCAP Vision Zero is something we feel strongly about, and our technology helps implement this vision in everyday cars. Security is another field where AI is creating enormous value. It might not be a revolutionary leap like in Automotive, but AI is penetrating in a big way. AI for surveillance and security operations – private, commercial and governmental – are all looking for better quality and efficiency gains. You don’t want to miss a single intruder, tailgater or emergency situation, and you certainly don’t want to spend your whole day going through false alarm notifications. You want to be able to react and even prevent bad things from happening. That is where AI comes in and people’s appetite for it is just getting bigger. Nowadays, they don’t just want the camera and the neural network to be there, they have an expectation of how they should perform: real-time, high-accuracy and for a wide coverage area. They require significant... --- - Published: 2021-09-22 - Modified: 2024-11-05 - URL: https://hailo.ai/zh-hans/blog/5-questions-on-the-edge-ai-with-hailo-ceo-orr-danon/ - Categories: AI Software, Edge AI Device, Intelligent Camera, Power Efficiency, Surveillance - Translation Priorities: 可选 Where do you see high-performance Edge AI making the most impact? Automotive autonomy and driver assistance (ADAS) immediately comes to mind. To make the roads safer we increasingly let technology take the wheel – whether it’s supporting drivers or replacing them altogether. This takes a lot of AI (Artificial Intelligence), which translates into a lot of AI processing power that needs to be very efficient and be spread throughout the entire vehicle, including the vehicle edge. We saw a natural fit for our architecture here, so we designed an automotive-grade chip. Euro NCAP Vision Zero is something we feel strongly about, and our technology helps implement this vision in everyday cars. Security is another field where AI is creating enormous value. It might not be a revolutionary leap like in Automotive, but AI is penetrating in a big way. AI for surveillance and security operations – private, commercial and governmental – are all looking for better quality and efficiency gains. You don’t want to miss a single intruder, tailgater or emergency situation, and you certainly don’t want to spend your whole day going through false alarm notifications. You want to be able to react and even prevent bad things from happening. That is where AI comes in and people’s appetite for it is just getting bigger. Nowadays, they don’t just want the camera and the neural network to be there, they have an expectation of how they should perform: real-time, high-accuracy and for a wide coverage area. They require significant... --- - Published: 2021-09-22 - Modified: 2024-11-05 - URL: https://hailo.ai/ja/blog/5-questions-on-the-edge-ai-with-hailo-ceo-orr-danon/ - Categories: AI Software, Edge AI Device, Intelligent Camera, Power Efficiency, Surveillance - Translation Priorities: 可选 Where do you see high-performance Edge AI making the most impact? Automotive autonomy and driver assistance (ADAS) immediately comes to mind. To make the roads safer we increasingly let technology take the wheel – whether it’s supporting drivers or replacing them altogether. This takes a lot of AI (Artificial Intelligence), which translates into a lot of AI processing power that needs to be very efficient and be spread throughout the entire vehicle, including the vehicle edge. We saw a natural fit for our architecture here, so we designed an automotive-grade chip. Euro NCAP Vision Zero is something we feel strongly about, and our technology helps implement this vision in everyday cars. Security is another field where AI is creating enormous value. It might not be a revolutionary leap like in Automotive, but AI is penetrating in a big way. AI for surveillance and security operations – private, commercial and governmental – are all looking for better quality and efficiency gains. You don’t want to miss a single intruder, tailgater or emergency situation, and you certainly don’t want to spend your whole day going through false alarm notifications. You want to be able to react and even prevent bad things from happening. That is where AI comes in and people’s appetite for it is just getting bigger. Nowadays, they don’t just want the camera and the neural network to be there, they have an expectation of how they should perform: real-time, high-accuracy and for a wide coverage area. They require significant... --- > This AI case study explores the shift from AI/ML environment to real world app deployment on embedded edge AI device - Published: 2021-09-01 - Modified: 2024-11-05 - URL: https://hailo.ai/blog/customer-case-study-developing-a-high-performance-application-on-an-embedded-edge-ai-device/ - Categories: AI Software, Edge AI Developer, Edge AI Device, Object Detection, TAPPAS - Translation Priorities: Optional One of the challenges in building embedded AI applications is taking it from machine learning research environment all the way to a full application deployed on an embedded host in the real world. Recently, a Hailo customer has asked us to bring up such an application. Our Applications Team thought it was a good opportunity to share the development process and its outcome as an AI case study and a typical example of advanced edge AI development. The Embedded AI Scenario Our customer required real-time high-accuracy object detection to run on a single video stream using an embedded host. The required input video resolution was HD (high definition, 720p). The chosen platform for this project is based on NXP’s i. MX8M ARM processor. The Hailo-8 AI processor is connected to it as an AI accelerator. Figure 1: The application's setup The object detection model that was selected for this application is YOLOv5m.  This modern object detection deep learning model allows both high detection accuracy and high throughput when running on the Hailo-8 device. It reaches an accuracy of 41. 7mAP on the COCO validation set. The requirements posed by our customer are by no means an easy ask. As embedded devices have strict power, heat dissipation and space constraints (you can read more about those in our paper on edge AI power efficiency), traditional processors designed for them are small and not very powerful. In most cases, such processors simply can’t support large, compute- or parameter-intense neural network models, processing... --- > Diese KI-Fallstudie untersucht den Wandel von der KI/ML-Umgebung zur realen App-Bereitstellung auf eingebetteten Edge-KI-Geräten. - Published: 2021-09-01 - Modified: 2024-06-23 - URL: https://hailo.ai/de/blog/customer-case-study-developing-a-high-performance-application-on-an-embedded-edge-ai-device/ - Categories: AI Software, Edge AI Developer, Edge AI Developer, Edge AI Device, Edge AI Device, Object Detection, TAPPAS - Translation Priorities: Optional One of the challenges in building embedded AI applications is taking it from machine learning research environment all the way to a full application deployed on an embedded host in the real world. Recently, a Hailo customer has asked us to bring up such an application. Our Applications Team thought it was a good opportunity to share the development process and its outcome as an AI case study and a typical example of advanced edge AI development. The Embedded AI Scenario Our customer required real-time high-accuracy object detection to run on a single video stream using an embedded host. The required input video resolution was HD (high definition, 720p). The chosen platform for this project is based on NXP’s i. MX8M ARM processor. The Hailo-8 AI processor is connected to it as an AI accelerator. Figure 1: The application's setup The object detection model that was selected for this application is YOLOv5m.  This modern object detection deep learning model allows both high detection accuracy and high throughput when running on the Hailo-8 device. It reaches an accuracy of 41. 7mAP on the COCO validation set. The requirements posed by our customer are by no means an easy ask. As embedded devices have strict power, heat dissipation and space constraints (you can read more about those in our paper on edge AI power efficiency), traditional processors designed for them are small and not very powerful. In most cases, such processors simply can’t support large, compute- or parameter-intense neural network models, processing... --- - Published: 2021-09-01 - Modified: 2024-11-05 - URL: https://hailo.ai/zh-hans/blog/customer-case-study-developing-a-high-performance-application-on-an-embedded-edge-ai-device/ - Categories: AI Software, Edge AI Developer, Edge AI Device, Object Detection, TAPPAS - Translation Priorities: 可选 One of the challenges in building embedded AI applications is taking it from machine learning research environment all the way to a full application deployed on an embedded host in the real world. Recently, a Hailo customer has asked us to bring up such an application. Our Applications Team thought it was a good opportunity to share the development process and its outcome as an AI case study and a typical example of advanced edge AI development. The Embedded AI Scenario Our customer required real-time high-accuracy object detection to run on a single video stream using an embedded host. The required input video resolution was HD (high definition, 720p). The chosen platform for this project is based on NXP’s i. MX8M ARM processor. The Hailo-8 AI processor is connected to it as an AI accelerator. Figure 1: The application's setup The object detection model that was selected for this application is YOLOv5m.  This modern object detection deep learning model allows both high detection accuracy and high throughput when running on the Hailo-8 device. It reaches an accuracy of 41. 7mAP on the COCO validation set. The requirements posed by our customer are by no means an easy ask. As embedded devices have strict power, heat dissipation and space constraints (you can read more about those in our paper on edge AI power efficiency), traditional processors designed for them are small and not very powerful. In most cases, such processors simply can’t support large, compute- or parameter-intense neural network models, processing... --- - Published: 2021-09-01 - Modified: 2024-11-05 - URL: https://hailo.ai/ja/blog/customer-case-study-developing-a-high-performance-application-on-an-embedded-edge-ai-device/ - Categories: AI Software, Edge AI Developer, Edge AI Device, Object Detection, TAPPAS - Translation Priorities: 可选 One of the challenges in building embedded AI applications is taking it from machine learning research environment all the way to a full application deployed on an embedded host in the real world. Recently, a Hailo customer has asked us to bring up such an application. Our Applications Team thought it was a good opportunity to share the development process and its outcome as an AI case study and a typical example of advanced edge AI development. The Embedded AI Scenario Our customer required real-time high-accuracy object detection to run on a single video stream using an embedded host. The required input video resolution was HD (high definition, 720p). The chosen platform for this project is based on NXP’s i. MX8M ARM processor. The Hailo-8 AI processor is connected to it as an AI accelerator. Figure 1: The application's setup The object detection model that was selected for this application is YOLOv5m.  This modern object detection deep learning model allows both high detection accuracy and high throughput when running on the Hailo-8 device. It reaches an accuracy of 41. 7mAP on the COCO validation set. The requirements posed by our customer are by no means an easy ask. As embedded devices have strict power, heat dissipation and space constraints (you can read more about those in our paper on edge AI power efficiency), traditional processors designed for them are small and not very powerful. In most cases, such processors simply can’t support large, compute- or parameter-intense neural network models, processing... --- > Edge machine learning is changing the game. Dive into our blog to understand how tiles optimize squeeze and excite operations for efficiency. - Published: 2021-08-11 - Modified: 2024-11-05 - URL: https://hailo.ai/blog/edge-ml-deep-dive-why-you-should-use-tiles-in-squeeze-and-excite-operations/ - Categories: AI Software, Edge AI Developer, ML Deep Dive - Translation Priorities: Optional This blog post presents the Tiled Squeeze-and-Excite (TSE) - a method designed to improve the deployment efficiency of the Squeeze-and-Excite (SE) operator to dataflow architecture AI accelerators like the Hailo-8 AI processor. It is based on our latest paper, which has recently been accepted by the 2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW) for the NeurArch Workshop! We developed TSE with the goal of accelerating the runtime of neural network architectures that include SE without requiring users to change or re-train their model. This work is part of a larger bag-of-tricks for model optimization prior to inference. Some of these methods have no effect on the model (e. g. , batch norm folding), some have an only numeric effect but preserve the architecture (e. g. , quantization) and some change the architecture to optimize performance. TSE falls in the latter category. What Is Squeeze-and-Excite? Squeeze-and-Excite (SE) (Hu, et al. , 2017), is a popular operation used in many deep learning networks to increase the accuracy with only a small increase to compute and parameters. The operation is part of a more general family of operations called channel attention. In channel attention we re-calibrate each channel of a given tensor according to the tensor’s data, which means that for every input we use different calibration factors. Unlike the convolution, which uses only a small receptive field, SE uses global context and operates in two steps. We first squeeze the input tensor by means of global average pooling (GAP). For an... --- - Published: 2021-08-11 - Modified: 2024-03-26 - URL: https://hailo.ai/de/blog/edge-ml-deep-dive-why-you-should-use-tiles-in-squeeze-and-excite-operations/ - Categories: AI Software, Edge AI Developer, Edge AI Developer, ML Deep Dive - Translation Priorities: Optional This blog post presents the Tiled Squeeze-and-Excite (TSE) - a method designed to improve the deployment efficiency of the Squeeze-and-Excite (SE) operator to dataflow architecture AI accelerators like the Hailo-8 AI processor. It is based on our latest paper, which has recently been accepted by the 2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW) for the NeurArch Workshop! We developed TSE with the goal of accelerating the runtime of neural network architectures that include SE without requiring users to change or re-train their model. This work is part of a larger bag-of-tricks for model optimization prior to inference. Some of these methods have no effect on the model (e. g. , batch norm folding), some have an only numeric effect but preserve the architecture (e. g. , quantization) and some change the architecture to optimize performance. TSE falls in the latter category. What Is Squeeze-and-Excite? Squeeze-and-Excite (SE) (Hu, et al. , 2017), is a popular operation used in many deep learning networks to increase the accuracy with only a small increase to compute and parameters. The operation is part of a more general family of operations called channel attention. In channel attention we re-calibrate each channel of a given tensor according to the tensor’s data, which means that for every input we use different calibration factors. Unlike the convolution, which uses only a small receptive field, SE uses global context and operates in two steps. We first squeeze the input tensor by means of global average pooling (GAP). For an... --- - Published: 2021-08-11 - Modified: 2024-11-05 - URL: https://hailo.ai/zh-hans/blog/edge-ml-deep-dive-why-you-should-use-tiles-in-squeeze-and-excite-operations/ - Categories: AI Software, Edge AI Developer, ML Deep Dive - Translation Priorities: 可选 This blog post presents the Tiled Squeeze-and-Excite (TSE) - a method designed to improve the deployment efficiency of the Squeeze-and-Excite (SE) operator to dataflow architecture AI accelerators like the Hailo-8 AI processor. It is based on our latest paper, which has recently been accepted by the 2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW) for the NeurArch Workshop! We developed TSE with the goal of accelerating the runtime of neural network architectures that include SE without requiring users to change or re-train their model. This work is part of a larger bag-of-tricks for model optimization prior to inference. Some of these methods have no effect on the model (e. g. , batch norm folding), some have an only numeric effect but preserve the architecture (e. g. , quantization) and some change the architecture to optimize performance. TSE falls in the latter category. What Is Squeeze-and-Excite? Squeeze-and-Excite (SE) (Hu, et al. , 2017), is a popular operation used in many deep learning networks to increase the accuracy with only a small increase to compute and parameters. The operation is part of a more general family of operations called channel attention. In channel attention we re-calibrate each channel of a given tensor according to the tensor’s data, which means that for every input we use different calibration factors. Unlike the convolution, which uses only a small receptive field, SE uses global context and operates in two steps. We first squeeze the input tensor by means of global average pooling (GAP). For an... --- - Published: 2021-08-11 - Modified: 2024-11-05 - URL: https://hailo.ai/ja/blog/edge-ml-deep-dive-why-you-should-use-tiles-in-squeeze-and-excite-operations/ - Categories: AI Software, Edge AI Developer, ML Deep Dive - Translation Priorities: 可选 This blog post presents the Tiled Squeeze-and-Excite (TSE) - a method designed to improve the deployment efficiency of the Squeeze-and-Excite (SE) operator to dataflow architecture AI accelerators like the Hailo-8 AI processor. It is based on our latest paper, which has recently been accepted by the 2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW) for the NeurArch Workshop! We developed TSE with the goal of accelerating the runtime of neural network architectures that include SE without requiring users to change or re-train their model. This work is part of a larger bag-of-tricks for model optimization prior to inference. Some of these methods have no effect on the model (e. g. , batch norm folding), some have an only numeric effect but preserve the architecture (e. g. , quantization) and some change the architecture to optimize performance. TSE falls in the latter category. What Is Squeeze-and-Excite? Squeeze-and-Excite (SE) (Hu, et al. , 2017), is a popular operation used in many deep learning networks to increase the accuracy with only a small increase to compute and parameters. The operation is part of a more general family of operations called channel attention. In channel attention we re-calibrate each channel of a given tensor according to the tensor’s data, which means that for every input we use different calibration factors. Unlike the convolution, which uses only a small receptive field, SE uses global context and operates in two steps. We first squeeze the input tensor by means of global average pooling (GAP). For an... --- > Empower your Edge AI development with Hailo's comprehensive overview of AI model zoos, including TensorFlow and ONNX ecosystems. - Published: 2021-07-20 - Modified: 2024-11-12 - URL: https://hailo.ai/blog/how-an-open-model-zoo-can-boost-your-edge-ai-system-development/ - Categories: AI Software, Edge AI Developer, TAPPAS - Translation Priorities: Optional Edge AI system development has been propelled forward by the exciting emergence of new purpose-built processing architectures that empower us to do so much more on the edge. An important, inseparable part of an edge AI offering is the dedicated software that enables the hardware. Of course, every new chip comes with its own software stack and development environment, but the extent those enable developers differ. In order to truly support developers, AI software needs not only to be robust but to give the developer flexibility and speed up Time-to-Market with an envelope of developer tools. An open AI Model Zoo is a great resource for developers, and a key part of this envelope. What is a Model Zoo? The Model Zoo is a repository which includes a wide range of pre-trained, pre-compiled deep learning NN models and tasks available in TensorFlow and ONNX formats. Its aim is to provide developers with quick functionality to hit the ground running by providing a binary HEF file per each pre-trained model and supporting with the Hailo toolchain and Application suite. With the goal of making it as easy as possible for developers to get up and running with the Hailo AI processors, we have built our own open Hailo Model Zoo repository, which includes: Pre-trained models – a large selection, demonstrating both the versatility and high AI performance of the Hailo-AI processors Pre-configured build flows – validated and optimized flows to take each network from an ONNX/TensorFlow model to a deployable binary, incorporating models evaluation and... --- > Stärken Sie Ihre Edge-KI-Entwicklung mit Hailos umfassendem Überblick über KI-Modellzoos, einschließlich TensorFlow- und ONNX-Ökosystemen. - Published: 2021-07-20 - Modified: 2024-11-12 - URL: https://hailo.ai/de/blog/how-an-open-model-zoo-can-boost-your-edge-ai-system-development/ - Categories: AI Software, AI Software, Edge AI Developer, Edge AI Developer, TAPPAS - Translation Priorities: Optional Die Entwicklung von Edge-KI-Systemen wurde durch die spannende Entstehung neuer speziell entwickelter Verarbeitungsarchitekturen vorangetrieben, die es uns ermöglichen, am Edge so viel mehr zu tun. Ein wichtiger, untrennbarer Bestandteil eines Edge-AI-Angebots ist die dedizierte Software, die die Hardware ermöglicht. Natürlich bringt jeder neue Chip seinen eigenen Software-Stack und seine eigene Entwicklungsumgebung mit, doch der Umfang dieser Möglichkeiten für Entwickler ist unterschiedlich. Um Entwickler wirklich zu unterstützen, muss KI-Software nicht nur robust sein, sondern dem Entwickler auch Flexibilität bieten und die Markteinführungszeit mithilfe zahlreicher Entwicklertools verkürzen. Ein offener AI Model Zoo ist eine großartige Ressource für Entwickler und ein wichtiger Teil dieses Rahmens. Was ist ein Modell Zoo? Der Model Zoo ist ein Repository, das eine große Palette vortrainierter, vorkompilierter Deep-Learning-NN-Modelle und -Aufgaben enthält, die in den Formaten TensorFlow und ONNX verfügbar sind. Ziel ist es, Entwicklern eine schnelle Funktionalität für den sofortigen Einstieg zu bieten, indem für jedes vortrainierte Modell eine binäre HEF-Datei bereitgestellt und die Hailo-Toolchain und Anwendungssuite unterstützt wird. Mit dem Ziel, Entwicklern den Einstieg in die Hailo-KI-Prozessoren, so einfach wie möglich zu machen, haben wir unser eigenes offenes Hailo Modell Zoo Repository erstellt, das Folgendes umfasst: Vortrainierte Modelle – eine große Auswahl, die sowohl die Vielseitigkeit als auch die hohe KI-Leistung der Hailo-AI-Prozessoren demonstriert Vorkonfigurierte Build-Flows – validierte und optimierte Flows, um jedes Netzwerk von einem ONNX/TensorFlow-Modell in eine bereitstellbare Binärdatei zu bringen, einschließlich Modellbewertung und Leistungsanalyse. Um eine einfache Auswertung zu ermöglichen, haben wir in unseren Model Zoo beliebte Modelle aus Open-Source-Repositories ohne Änderungen aufgenommen und mit öffentlich verfügbaren Datensätzen trainiert. Außerdem wurde... --- > 利用 Hailo 对 AI 模型库（包括 TensorFlow 和 ONNX 生态系统）的全面概述来增强您的边缘 AI 开发。 - Published: 2021-07-20 - Modified: 2024-12-24 - URL: https://hailo.ai/zh-hans/blog/how-an-open-model-zoo-can-boost-your-edge-ai-system-development/ - Categories: AI Software, Edge AI Developer, TAPPAS - Translation Priorities: 可选 振奋人心的新型专用处理架构的出现推动了边缘AI系统的发展，这些架构使我们能够在边缘做更多事情。边缘AI产品的一个不可分割的重要部分是支持硬件的专用软件。当然，每个新芯片都有自己的软件栈和开发环境，但对开发者的支持程度有差异 为了真正支持开发者，AI软件不仅需要稳健，还需要赋予开发者灵活性，通过一系列开发者工具加快产品上市时间。开放式人工智能Model Zoo对开发者来说是一个很好的资源，也是这个包络的关键部分。 什么是Model Zoo？ Model Zoo是一个存储库，其中包含广泛的预训练、预编译深度学习NN模型和任务，以TensorFlow和ONNX格式提供。其目标是通过为每个预训练模型提供二进制HEF文件，并支持Hailo工具链和应用套件，为开发者提供快速功能。 为了使开发者能够尽可能轻松地使用Hailo AI处理器，我们构建了自己的开放式Hailo Model Zoo，其中包括： 预训练模型 – 一个大型选择集，展示了Hailo-AI处理器的多功能性和高人工智能性能。 预配置构建流 – 经过验证和优化的流程，将每个网络从ONNX/TensorFlow模型转化为可部署的二进制模型，并纳入模型评估和性能分析。 为了实现简单的评估，我们在Model Zoo中包含了从开源存储库中获取的热门模型，这些模型未经修改，并在公开可用的数据集上进行了训练。此外还添加了到模型源链接，让用户可以调整模型，例如通过在他们自己的自定义数据集上进行训练。 此外，为了帮助用户为他们的AI应用选择最佳的神经网络模型，Hailo Model Explorer可以帮助用户做出更明智的决策，确保Hailo平台的最大效率。Model Explorer提供一个交互式界面，其中包含基于Hailo设备、任务、模型、FPS和准确性的过滤器，让用户可以从Hailo的庞大库中探索大量神经网络模型。 图1：Hailo Model Zoo – 从ONNX/TensorFlow格式的开源预训练模型到使用优化构建流编译的部署就绪产品 如何运行 Model Zoo采用Hailo数据流编译器，实现从预训练模型（ckpt/ONNX）到可以在Hailo-8上执行的最终Hailo可执行格式（HEF）的完整流程。为此，Hailo Model Zoo为用户提供以下功能： 解析：将Tensorflow/ONNX模型转换为Hailo的内部表示，其中包括模型的网络拓扑和原始权重。这个阶段的输出是一个HailoArchive文件（HAR）。 概要: 一份报告，其中包含预期模型在Hailo-8上的性能，包括FPS、延迟、功耗和HTML格式的各层完整分解。 量化：通过将其从全精度转换为有限整数位精度（4/8/16）来优化运行时模型，同时最小化精度退化。这个阶段包括几个算法，通过保证量化模型的准确性来优化性能。此阶段的输出是包含量化权重的Hailo Archive文件。 评估：在常用数据集（如ImageNet、COCO）上评估模型的准确性。使用我们的数值仿真器或Hailo-8可以对全精度模型和量化模型进行评估 编译：编译量化模型，生成可部署在Hailo-8芯片上的Hailo可执行格式（HEF）文件。 图2：Hailo Model Zoo功能流程图。蓝色是使用Hailo数据流编译器的模块 Hailo Model Zoo的最后阶段生成用于最终应用的HEF文件。举例来说，这是ADAS应用YOLOv3对象检测网络的HEF文件或智能家居相机解决方案CenterPose（姿态估计）网络的HEF文件生成的地方。 选择了我们存储库中的模型，以涵盖各种各样的通用架构和任务。它热门的先进架构，如YOLOv3、YOLOv4、CenterPose、CenterNet、ResNet-50等，其中许多已经是我们的AI基准的一部分。为了实现轻松的基准测试，每个预训练模型都有自己的预处理/后处理功能和数据集采集流程，以使评估可行。我们将所有这些功能打包到一个易于使用的包中，因此它可以简单快速地复制，并且可以在广泛的网络中测量性能。借助Hailo-8 AI处理器强大的AI加速功能，用户可以利用Model Zoo在边缘设备中为这些热门模型提供卓越的性能。 这与TAPPAS有什么关系 用户可以将AI Model Zoo中的加速模型集成到自己的应用中。将Hailo Model Zoo与我们的TAPPAS高性能应用工具包结合，边缘AI开发人员可以在短时间内构建和部署有意义的应用。事实上，TAPPAS中的所有应用都建立在Model Zoo网络之上 以一个常见的用例为例，姿态估计可以获得有关购物者行为和愿望的有意义洞察，用于智慧零售智能视频分析（IVA）。开发这种应用的用户可能会发现基于预训练的姿态估计模型（如我们的model-zoo中的CenterPose_RegNetX_1. 6GF）来开发很有用。基于Hailo TAPPAS，模型可以轻松集成到您的应用中。 图3：我们的预训练姿态估计模型在仓库安全摄像头镜头上的Hailo-8 AI处理器上运行 另一个有趣的用例是智慧城市应用的智能交通监控。您可以使用现有的预编译模型SSD_MobileNet_v1 VisDrone立即启动并运行应用。事实上，Hailo的一个小型工程师团队在Hailo首次黑客马拉松中24小时就做到了！ 图4：Hailo-8 AI处理器在繁忙城市高速公路的高分辨率实时视频源上运行作为TAPPAS Tiling应用一部分的SSD_MobileNet_v1_VisDrone模型（在此处查看我们的完整演示）） Hailo Model Zoo为边缘人工智能开发者提供一个综合环境，以生成他们的下一个深度学习应用，为各种有用的神经任务提供开箱即用的预训练模型和应用解决方案。展望未来，我们将不断添加新的先进模型，涵盖新的任务和元架构，改进我们的预配置构建流和优化流程，以帮助开发者构建卓越的边缘AI应用，充分利用Hailo-8 AI处理器的行业领先功能。 如需了解更多信息，请查看我们的 或访问Hailo Model Zoo Github --- > TensorFlow や ONNX エコシステムを含む、Hailo の AI モデル ズーの包括的な概要を活用して、エッジ AI 開発を強化します。 - Published: 2021-07-20 - Modified: 2024-12-24 - URL: https://hailo.ai/ja/blog/how-an-open-model-zoo-can-boost-your-edge-ai-system-development/ - Categories: AI Software, Edge AI Developer, TAPPAS - Translation Priorities: 可选 エッジAIシステムの開発は、エッジ上でより多くのことができるようにする新しい目的別建築処理のエキサイティングな出現によって推進されています。エッジAIの重要な不可分の部分は、ハードウェアを可能にする専用ソフトウェアです。もちろん、どの新型チップにも独自のソフトウェア・スタックと開発環境が付属しているが、それらが開発者をどの程度サポートするかは異なります。 開発者を真にサポートするためには、AIソフトウェアは堅牢であるだけでなく、開発者に柔軟性を与え、開発者向けツールのエンベロープによって市場投入までの時間を短縮する必要です。オープンなAI Model Zooは、開発者にとって素晴らしいリソースであり、このエンベロープの重要な部分です Model Zooとはなんでしょうか？ Model Zoo は、TensorFlowとONNX形式で利用可能な、事前に研修され、コンパイル済みの深層学習NNモデルとタスクを幅広く含むリポジトリです。その目的は、各訓練済みモデルごとにバイナリのHEFファイルを提供し、Hailoツールチェーンとアプリ・スイートをサポートすることで、開発者に素早く実行に移せる機能を提供することです。 開発者ができるだけ簡単に Hailo AI プロセッサ を使用できるようにすることを目標に、当社は独自のオープンな Hailo Model Zoo リポジトリを構築しました： 事前研修済モデル - Hailo-AIプロセッサの多用途性とAI性能の高さを示す豊富な選択肢 事前設定された構築フロー - ONNX/TensorFlowモデルから展開可能なバイナリに各ネットワークを移行するための検証および最適化されたフローで、モデルの評価とパフォーマンス分析が組み込まれています。 簡単な評価を可能にするために、オープンソースのリポジトリから取得した一般的なモデルを修正することなくモデルズーに含め、一般に利用可能なデータセットで研修させた。また、ユーザーが自分のカスタムデータセットでモデルを研修するなど、モデルを調整したい場合に備えて、モデルソースへのリンクも追加された。 さらに、ユーザーがAIアプリに最適なNNモデルを選択できるように、Hailo Model Explorerは、より良い情報に基づいた意思決定を支援し、Hailoプラットフォームでの最大限の効率を保証します。Model Explorerは、Hailo デバイス、タスク、モデル、FPS、精度に基づくフィルターを備えた対話型インターフェイスを提供し、ユーザーが Hailo の膨大なライブラリから多数の NN モデルを探索できるようにします。 図1：Hailo Model Zoo – ONNX/TensorFlowで事前研修されたオープンソースのモデルから、最適化されたビルドフローを使用してコンパイルされたデプロイ可能な製品です。 働き方 Model Zooは、Hailo データフローコンパイラ を使用して、事前に研修されたモデル (ckpt/ONNX) から、Hailo-8で実行可能な最終的なHailo実行形式 (HEF) までの完全なフローを作成します。そのために、Hailo Model Zooは以下の機能をユーザーに提供する： 解析：Tensorflow/ONNXモデルをHailoの内部表現に変換します。この内部表現には、ネットワークのトポロジーとモデルの元の重みが含まれます。このステージの出力は、Hailo Archive ファイル (HAR) です。 プロファイル：Hailo-8で期待されるモデルのパフォーマンス（FPS、レイテンシー、消費電力、各レイヤーの完全な内訳を含む）をHTML形式でまとめたレポート。 量子化：精度の劣化を最小限に抑えながら、モデルを全精度から限られた整数ビット精度（4/8/16）に変換し、実行時に最適化する。この段階には、量子化モデルの精度を保証することで性能を最適化するいくつかのアルゴリズムが含まれる。このステージの出力は、量子化された重みを含むHailo Archiveファイルです。 評価：一般的なデータセット（ImageNet、COCOなど）でモデルの精度を評価する。評価は、数値エミュレータまたはHailo-8を使用して、全精度モデルと量子化モデルで行うことができます。 コンパイル: 量子化されたモデルをコンパイルして、Hailo-8チップ上に配置できるHailo実行形式(HEF)ファイルを生成します。 図2：Hailo Model Zooの機能フローチャート。青字は、Hailo データフローコンパイラを使用するブロックです。 Hailo Model Zooの最終段階は、最終アプリで使用するHEFファイルを生成します。例えるなら、ADASアプリ用のYOLOv3オブジェクト検出ネットワークのHEFファイルや、スマートホームカメラソリューション用のCenterPose（姿勢推定）ネットワークのHEFファイルが生成される場所です。 当社のリポジトリにあるモデルは、一般的なアーキテクチャやタスクを幅広くカバーするように選択されております。YOLOv3、YOLOv4、CenterPose、CenterNet、ResNet-50など、人気の高い最先端のアーキテクチャが含まれており、その多くはすでに我々のAIベンチマークの一部となっている。簡単なベンチマークを可能にするため、すべての事前研修済モデルには、独自の事前/事後処理関数とデータセット取得フローが付属しており、評価の実現性を高めています。私たちは、これらの機能をすべて使いやすいパッケージにまとめました。そのため、再現が簡単で速く、幅広いネットワークでパフォーマンスを測定することができます。Hailo-8 AIプロセッサーの強力なAIアクセラレーション機能により、ユーザーはModel Zooを活用して、エッジデバイスのこれらの人気モデルで卓越した性能を達成することができます。 TAPPASとの関係 ユーザーは、AI Model Zooの加速モデルを自分のアプリに統合することが可能です。Hailo Model Zooを当社のTAPPAS 高性能アプリツールキットと組み合わせて使用することで、エッジAI開発者は有意義なアプリケーションを短期間で構築し、展開することができます。実際、TAPPASのすべてのアプリは、Model Zooネットワークの上に構築されています。 一般的なユースケースを例に挙げると、ポーズ推定によって買い物客の行動や欲求に関する有意義な洞察を得ることができ、スマートリテールで使用できます。インテリジェントビデオアナリティクス（IVA）に対してこの種のアプリを開発するユーザーは、モデルズーにあるCenterPose_RegNetX_1. 6GFのような事前に研修されたポーズ推定モデルに基づいて開発すると便利です。Hailo TAPPASをベースにしたこのモデルは、お客様のアプリケーションに簡単に統合することができます。 図3：Hailo-8AIプロセッサーで倉庫の監視カメラ映像を処理する、事前に訓練されたポーズ推定モデル もう一つの興味深い使用例は、スマートシティアプリケーションのためのインテリジェントな交通監視である。既存のコンパイル済みモデル SSD_MobileNet_v1 VisDrone を使用して、すぐにアプリを立ち上げて実行できます。実際、Hailoのエンジニアの小さなチームは、Hailoの最初のハッカソンで、24時間でそれを成し遂げた！ 図4: TAPPASタイリング・アプリケーションの一部としてSSD_MobileNet_v1_VisDroneモデルを実行するHailo-8 AIプロセッサと、交通量の多い都市高速道路の高解像度ライブ・ビデオ・フィード（デモの全容はこちらを参照してください）。 Hailo Model Zooは、エッジAI開発者が次の深層学習アプリを生成するための包括的な環境を提供し、すぐに使える事前研修済モデルと、さまざまな有用なニューラル・タスク用のアプリ・ソリューションを備えています。今後、Hailo-8 AI Processorの業界をリードする機能を活用し、開発者が可能な限り最高のエッジAIアプリを構築できるよう、新しい最先端のモデルを継続的に追加し、新しいタスクやメタアーキテクチャをカバーし、設定済みのビルドフローや最適化プロセスを改善していきます。 詳しくは、Hailo Model Zoo をご覧いただくか、Hailo Model Zoo Github をご覧ください。 --- > See how smart retail solutions blend technology and retail to optimize shopping experiences. AI vision redefines the customer experience. - Published: 2021-07-01 - Modified: 2024-11-05 - URL: https://hailo.ai/blog/the-edge-of-retail-why-powerful-intelligent-vision-is-the-future-of-brick-mortar-stores/ - Categories: AI Hardware, Edge AI Device, Intelligent Camera, Smart Retail - Translation Priorities: Optional Brick-and-mortar retail has been facing tough competition from eCommerce, with the COVID-19 tailwind reinforcing the growing transition from brick-and-mortar to eCommerce. Smart Retail solutions and technologies are helping physical stores compete against online ones by leveling the playing field. Digitalizing and deploying AI for automation and advanced analytics creates a physical shopping experience that is on par, if not better, with the entirely digital one, with benefits both to the customer and the retailer. The customer enjoys an improved in-store experience – a walk through a streamlined, easy to navigate store, a no-queue "Just Walk Out" automated checkout and personalized offers and functions. It is very similar to shopping online, but with the added benefits of being able to see and touch the products. The experience may also be a multi-channel one, in which customers can choose between shopping in-store and ordering a delivery or in-store pickup and connect to online resources while in-store for navigation, additional information, customization, digital try-on and more. For the retailer, intelligent automation and in-store analytics translate into operational optimization and cost savings. AI is used to increase customer traffic, improve loss prevention measures, and optimize store layout, shelves and displays, inventory and personnel work. It provides the data and insight that are used, among other things, to increase footfall, better target promotions and nurture long-term relationships with customers. In Smart Retail, Intelligent Vision Is King While there are many types of sensors used in Smart Retail and IoT solutions (e. g. , RF, motion,... --- > Erfahren Sie, wie intelligente Einzelhandelslösungen Technologie und Einzelhandel vereinen, um das Einkaufserlebnis zu optimieren. - Published: 2021-07-01 - Modified: 2024-06-23 - URL: https://hailo.ai/de/blog/the-edge-of-retail-why-powerful-intelligent-vision-is-the-future-of-brick-mortar-stores/ - Categories: AI Hardware, AI Hardware, Edge AI Device, Edge AI Device, Intelligent Camera, Intelligent Camera, Smart Retail - Translation Priorities: Optional Brick-and-mortar retail has been facing tough competition from eCommerce, with the COVID-19 tailwind reinforcing the growing transition from brick-and-mortar to eCommerce. Smart Retail technologies are helping physical stores compete against online ones by leveling the playing field. Digitalizing and deploying AI for automation and advanced analytics creates a physical shopping experience that is on par, if not better, with the entirely digital one, with benefits both to the customer and the retailer. The customer enjoys an improved in-store experience – a walk through a streamlined, easy to navigate store, a no-queue "Just Walk Out" automated checkout and personalized offers and functions. It is very similar to shopping online, but with the added benefits of being able to see and touch the products. The experience may also be a multi-channel one, in which customers can choose between shopping in-store and ordering a delivery or in-store pickup and connect to online resources while in-store for navigation, additional information, customization, digital try-on and more. For the retailer, intelligent automation and in-store analytics translate into operational optimization and cost savings. AI is used to increase customer traffic, improve loss prevention measures, and optimize store layout, shelves and displays, inventory and personnel work. It provides the data and insight that are used, among other things, to increase footfall, better target promotions and nurture long-term relationships with customers. In Smart Retail, Intelligent Vision Is King While there are many types of sensors used in Smart Retail and IoT solutions (e. g. , RF, motion, pressure, temperature... --- - Published: 2021-07-01 - Modified: 2024-11-05 - URL: https://hailo.ai/zh-hans/blog/the-edge-of-retail-why-powerful-intelligent-vision-is-the-future-of-brick-mortar-stores/ - Categories: AI Hardware, Edge AI Device, Intelligent Camera, Smart Retail - Translation Priorities: 可选 Brick-and-mortar retail has been facing tough competition from eCommerce, with the COVID-19 tailwind reinforcing the growing transition from brick-and-mortar to eCommerce. Smart Retail solutions and technologies are helping physical stores compete against online ones by leveling the playing field. Digitalizing and deploying AI for automation and advanced analytics creates a physical shopping experience that is on par, if not better, with the entirely digital one, with benefits both to the customer and the retailer. The customer enjoys an improved in-store experience – a walk through a streamlined, easy to navigate store, a no-queue "Just Walk Out" automated checkout and personalized offers and functions. It is very similar to shopping online, but with the added benefits of being able to see and touch the products. The experience may also be a multi-channel one, in which customers can choose between shopping in-store and ordering a delivery or in-store pickup and connect to online resources while in-store for navigation, additional information, customization, digital try-on and more. For the retailer, intelligent automation and in-store analytics translate into operational optimization and cost savings. AI is used to increase customer traffic, improve loss prevention measures, and optimize store layout, shelves and displays, inventory and personnel work. It provides the data and insight that are used, among other things, to increase footfall, better target promotions and nurture long-term relationships with customers. In Smart Retail, Intelligent Vision Is King While there are many types of sensors used in Smart Retail and IoT solutions (e. g. , RF, motion,... --- - Published: 2021-07-01 - Modified: 2024-11-05 - URL: https://hailo.ai/ja/blog/the-edge-of-retail-why-powerful-intelligent-vision-is-the-future-of-brick-mortar-stores/ - Categories: AI Hardware, Edge AI Device, Intelligent Camera, Smart Retail - Translation Priorities: 可选 Brick-and-mortar retail has been facing tough competition from eCommerce, with the COVID-19 tailwind reinforcing the growing transition from brick-and-mortar to eCommerce. Smart Retail solutions and technologies are helping physical stores compete against online ones by leveling the playing field. Digitalizing and deploying AI for automation and advanced analytics creates a physical shopping experience that is on par, if not better, with the entirely digital one, with benefits both to the customer and the retailer. The customer enjoys an improved in-store experience – a walk through a streamlined, easy to navigate store, a no-queue "Just Walk Out" automated checkout and personalized offers and functions. It is very similar to shopping online, but with the added benefits of being able to see and touch the products. The experience may also be a multi-channel one, in which customers can choose between shopping in-store and ordering a delivery or in-store pickup and connect to online resources while in-store for navigation, additional information, customization, digital try-on and more. For the retailer, intelligent automation and in-store analytics translate into operational optimization and cost savings. AI is used to increase customer traffic, improve loss prevention measures, and optimize store layout, shelves and displays, inventory and personnel work. It provides the data and insight that are used, among other things, to increase footfall, better target promotions and nurture long-term relationships with customers. In Smart Retail, Intelligent Vision Is King While there are many types of sensors used in Smart Retail and IoT solutions (e. g. , RF, motion,... --- > Hailo delves into the implications of deploying edge AI co-processors, offering insights into achieving high-performance AI at the edge. - Published: 2021-06-17 - Modified: 2025-03-24 - URL: https://hailo.ai/blog/mind-the-system-gap-system-level-implications-for-high-performance-edge-ai-coprocessors/ - Categories: AI Benchmarks, AI Hardware, Edge AI Box, Edge AI Developer - Translation Priorities: Optional Intro The new generation of domain-specific AI computing architectures is booming. The need for these and the benefit they provide to end applications are apparent. However, the inherent challenges in their development and development with them may not be. Specifically, the out-of-the-box interoperability and ease-of-use of decades-old architectures is the expected user and developer experience, but it does not go without saying when you build a new high performance processor from scratch, especially one with a novel architecture. These are validation and system challenges. The AI processor needs to deliver the best performance within a system’s constraints and without burdening its resources. It needs to integrate into common ecosystems. A novel architecture amplifies this inherent challenge because there is little existing legacy to rely on. Entering a new architectural domain requires new hardware and compatible software to be designed (you can read more about it in our domain-specific architectures post). You need to devise your own system benchmarking and validation processes and you will not have common-use feedback on their results. It is very common to evaluate an AI chip’s performance based on common industry benchmarks using parameters such as latency, throughput (frames or inferences per second) and power consumption. However, vendor-provided numbers need to be taken with a grain-of-salt. Measurements are done in laboratory conditions and neglect the effect the entire system will have on the accelerator’s performance and efficiency inside real-world edge devices. System validation of an AI accelerator is taking the evaluation beyond sterile lab conditions. The System... --- > Visit Hailo's blog to learn about validation and system challenges, as the AI processor needs to deliver the best performance within a system’s constraints and without burdening its resources. - Published: 2021-06-17 - Modified: 2024-10-09 - URL: https://hailo.ai/de/blog/mind-the-system-gap-system-level-implications-for-high-performance-edge-ai-coprocessors/ - Categories: AI Benchmarks, AI Hardware, AI Hardware, Edge AI Box, Edge AI Developer, Edge AI Developer - Translation Priorities: Optional Intro The new generation of domain-specific AI computing architectures is booming. The need for these and the benefit they provide to end applications are apparent. However, the inherent challenges in their development and development with them may not be. Specifically, the out-of-the-box interoperability and ease-of-use of decades-old architectures is the expected user and developer experience, but it does not go without saying when you build a new high performance processor from scratch, especially one with a novel architecture. These are validation and system challenges. The AI processor needs to deliver the best performance within a system’s constraints and without burdening its resources. It needs to integrate into common ecosystems. A novel architecture amplifies this inherent challenge because there is little existing legacy to rely on. Entering a new architectural domain requires new hardware and compatible software to be designed (you can read more about it in our domain-specific architectures post). You need to devise your own system benchmarking and validation processes and you will not have common-use feedback on their results. It is very common to evaluate an AI chip’s performance based on common industry benchmarks using parameters such as latency, throughput (frames or inferences per second) and power consumption. However, vendor-provided numbers need to be taken with a grain-of-salt. Measurements are done in laboratory conditions and neglect the effect the entire system will have on the accelerator’s performance and efficiency inside real-world edge devices. System validation of an AI accelerator is taking the evaluation beyond sterile lab conditions. The System... --- - Published: 2021-06-17 - Modified: 2025-03-24 - URL: https://hailo.ai/zh-hans/blog/mind-the-system-gap-system-level-implications-for-high-performance-edge-ai-coprocessors/ - Categories: AI Benchmarks, AI Hardware, Edge AI Box, Edge AI Developer - Translation Priorities: 可选 Intro The new generation of domain-specific AI computing architectures is booming. The need for these and the benefit they provide to end applications are apparent. However, the inherent challenges in their development and development with them may not be. Specifically, the out-of-the-box interoperability and ease-of-use of decades-old architectures is the expected user and developer experience, but it does not go without saying when you build a new high performance processor from scratch, especially one with a novel architecture. These are validation and system challenges. The AI processor needs to deliver the best performance within a system’s constraints and without burdening its resources. It needs to integrate into common ecosystems. A novel architecture amplifies this inherent challenge because there is little existing legacy to rely on. Entering a new architectural domain requires new hardware and compatible software to be designed (you can read more about it in our domain-specific architectures post). You need to devise your own system benchmarking and validation processes and you will not have common-use feedback on their results. It is very common to evaluate an AI chip’s performance based on common industry benchmarks using parameters such as latency, throughput (frames or inferences per second) and power consumption. However, vendor-provided numbers need to be taken with a grain-of-salt. Measurements are done in laboratory conditions and neglect the effect the entire system will have on the accelerator’s performance and efficiency inside real-world edge devices. System validation of an AI accelerator is taking the evaluation beyond sterile lab conditions. The System... --- - Published: 2021-06-17 - Modified: 2025-03-24 - URL: https://hailo.ai/ja/blog/mind-the-system-gap-system-level-implications-for-high-performance-edge-ai-coprocessors/ - Categories: AI Benchmarks, AI Hardware, Edge AI Box, Edge AI Developer - Translation Priorities: 可选 Intro The new generation of domain-specific AI computing architectures is booming. The need for these and the benefit they provide to end applications are apparent. However, the inherent challenges in their development and development with them may not be. Specifically, the out-of-the-box interoperability and ease-of-use of decades-old architectures is the expected user and developer experience, but it does not go without saying when you build a new high performance processor from scratch, especially one with a novel architecture. These are validation and system challenges. The AI processor needs to deliver the best performance within a system’s constraints and without burdening its resources. It needs to integrate into common ecosystems. A novel architecture amplifies this inherent challenge because there is little existing legacy to rely on. Entering a new architectural domain requires new hardware and compatible software to be designed (you can read more about it in our domain-specific architectures post). You need to devise your own system benchmarking and validation processes and you will not have common-use feedback on their results. It is very common to evaluate an AI chip’s performance based on common industry benchmarks using parameters such as latency, throughput (frames or inferences per second) and power consumption. However, vendor-provided numbers need to be taken with a grain-of-salt. Measurements are done in laboratory conditions and neglect the effect the entire system will have on the accelerator’s performance and efficiency inside real-world edge devices. System validation of an AI accelerator is taking the evaluation beyond sterile lab conditions. The System... --- > Hailo sheds light on the significance of TOPS AI and why it's crucial to look beyond this metric when assessing AI accelerator performance. - Published: 2021-06-02 - Modified: 2024-11-05 - URL: https://hailo.ai/blog/evaluating-edge-ai-accelerator-performance-why-tops-are-not-enough/ - Categories: AI Benchmarks, Edge AI Developer - Translation Priorities: Optional The recent requirement for efficient edge AI solutions across Smart City, Smart Retail, Surveillance and Automotive, has driven a boom in edge AI chips. Both established corporations and innovative startups have produced such offerings. When an edge AI solutions provider tries to evaluate the performance of the myriad chips available, he or she meets a single key metric: the tera operations per second, or TOPS. You do not have to look long before you encounter multiple such claims made by edge AI offerings. This is the way the market talks and we at Hailo are no exception, as our marketing and press releases show. What Are TOPS In AI? TOPS, in the way that the term is used here, is a measure of the maximum achievable AI performance of an AI accelerator or SoC. However, it is really a proxy for the physical properties of the device – the number of hardware Multiply-And-Accumulate (MAC) units implemented in the processor and the operating frequency. It can easily be calculated as follows: (The constant 2 is added because each MAC performs two operations: Multiply and Accumulate). The resulting number reflects the maximum achievable performance for a synthetic workload that does not realize any actual perception workload on the chip. Therefore, comparing AI accelerators based on this metric alone will not help uncover the performance KPIs that are interesting to the edge AI product designer – namely, throughput or latency. AI Processor Benchmarking and TOPs Utilization This fact has not gone unnoticed in... --- --- ## Resources - Published: 2023-07-31 - Modified: 2024-11-05 - URL: https://youtu.be/RtW1XLTum-A - Resource Industries: Security - Resource Types: Webinar Video Analytics re-imagined. The impact of powerful edge AI --- - Published: 2023-07-31 - Modified: 2024-11-20 - URL: https://youtu.be/RtW1XLTum-A - Resource Industries: Security - Resource Types: Webinar Video Analytics re-imagined. The impact of powerful edge AI --- - Published: 2023-07-31 - Modified: 2024-11-05 - URL: https://youtu.be/RtW1XLTum-A - Resource Industries: Security - Resource Types: Webinar Video Analytics re-imagined. The impact of powerful edge AI --- - Published: 2023-07-31 - Modified: 2024-11-05 - URL: https://youtu.be/RtW1XLTum-A - Resource Industries: Security - Resource Types: Webinar Video Analytics re-imagined. The impact of powerful edge AI --- - Published: 2023-03-26 - Modified: 2024-11-05 - URL: https://attendee.gotowebinar.com/register/8304429232497133069?source=Hailo+website - Resource Industries: Automotive - Resource Types: Webinar Joined Webinar --- - Published: 2023-03-26 - Modified: 2024-11-20 - URL: https://attendee.gotowebinar.com/register/8304429232497133069?source=Hailo+website - Resource Industries: Automotive - Resource Types: Webinar Joined Webinar --- - Published: 2023-03-26 - Modified: 2024-11-05 - URL: https://attendee.gotowebinar.com/register/8304429232497133069?source=Hailo+website - Resource Industries: Automotive - Resource Types: Webinar Joined Webinar --- - Published: 2023-03-26 - Modified: 2024-11-05 - URL: https://attendee.gotowebinar.com/register/8304429232497133069?source=Hailo+website - Resource Industries: Automotive - Resource Types: Webinar Joined Webinar --- - Published: 2022-11-03 - Modified: 2024-11-05 - URL: https://youtu.be/Ewe4tcKtX0Y - Resource Industries: Automotive - Resource Types: Webinar Scaling AI in Automotive with Hailo AI Acceleration --- - Published: 2022-11-03 - Modified: 2024-11-20 - URL: https://youtu.be/Ewe4tcKtX0Y - Resource Industries: Automotive - Resource Types: Webinar Scaling AI in Automotive with Hailo AI Acceleration --- - Published: 2022-11-03 - Modified: 2024-11-05 - URL: https://youtu.be/Ewe4tcKtX0Y - Resource Industries: Automotive - Resource Types: Webinar Scaling AI in Automotive with Hailo AI Acceleration --- - Published: 2022-11-03 - Modified: 2024-11-05 - URL: https://youtu.be/Ewe4tcKtX0Y - Resource Industries: Automotive - Resource Types: Webinar Scaling AI in Automotive with Hailo AI Acceleration --- - Published: 2021-08-26 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=jBlEXu5UOsY - Resource Industries: Automotive - Resource Types: Video Hailo for Automotive --- - Published: 2021-08-26 - Modified: 2024-11-20 - URL: https://www.youtube.com/watch?v=jBlEXu5UOsY - Resource Industries: Automotive - Resource Types: Video Hailo for Automotive --- - Published: 2021-08-26 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=jBlEXu5UOsY - Resource Industries: Automotive - Resource Types: Video Hailo for Automotive --- - Published: 2021-08-26 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=jBlEXu5UOsY - Resource Industries: Automotive - Resource Types: Video Hailo for Automotive --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2021-05-21 - Modified: 2024-11-06 - URL: https://youtu.be/aAaCX9Zcx8o - Resource Types: Presentations Linley Spring Processor Conference 2021 Daniel Chibotero, Chief Architect, Hailo: Scalable and Power-Efficient Architecture for Deep Learning at the Edge --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2021-05-21 - Modified: 2024-11-20 - URL: https://youtu.be/aAaCX9Zcx8o - Resource Types: Presentation Linley Spring Processor Conference 2021 Daniel Chibotero, Chief Architect, Hailo: Scalable and Power-Efficient Architecture for Deep Learning at the Edge --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2021-05-21 - Modified: 2024-11-06 - URL: https://youtu.be/aAaCX9Zcx8o - Resource Types: Presentations Linley Spring Processor Conference 2021 Daniel Chibotero, Chief Architect, Hailo: Scalable and Power-Efficient Architecture for Deep Learning at the Edge --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2021-05-21 - Modified: 2024-11-06 - URL: https://youtu.be/aAaCX9Zcx8o - Resource Types: Presentations Linley Spring Processor Conference 2021 Daniel Chibotero, Chief Architect, Hailo: Scalable and Power-Efficient Architecture for Deep Learning at the Edge --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2021-05-12 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=NUECoRCZoDE - Resource Types: Presentations Hailo Talk at PyCon Israel: Noy Nakash & Avi Naftalis: Collecting and Analyzing Data from PyTest with Elasticsearch --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2021-05-12 - Modified: 2024-11-20 - URL: https://www.youtube.com/watch?v=NUECoRCZoDE - Resource Types: Presentation Hailo Talk at PyCon Israel: Noy Nakash & Avi Naftalis: Collecting and Analyzing Data from PyTest with Elasticsearch --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2021-05-12 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=NUECoRCZoDE - Resource Types: Presentations Hailo Talk at PyCon Israel: Noy Nakash & Avi Naftalis: Collecting and Analyzing Data from PyTest with Elasticsearch --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2021-05-12 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=NUECoRCZoDE - Resource Types: Presentations Hailo Talk at PyCon Israel: Noy Nakash & Avi Naftalis: Collecting and Analyzing Data from PyTest with Elasticsearch --- - Published: 2021-03-21 - Modified: 2024-11-06 - URL: https://hailo.ai/wp-content/uploads/2023/10/Exploring-Neural-Networks-Quantizationvia-Layer-Wise-Quantization-Analysis.pdf - Resource Types: Article Recovering Neural Network Quantization Error Through Weight Factorization --- - Published: 2021-03-21 - Modified: 2024-11-20 - URL: https://hailo.ai/wp-content/uploads/2023/10/Exploring-Neural-Networks-Quantizationvia-Layer-Wise-Quantization-Analysis.pdf - Resource Types: Artikel Recovering Neural Network Quantization Error Through Weight Factorization --- - Published: 2021-03-21 - Modified: 2024-11-06 - URL: https://hailo.ai/wp-content/uploads/2023/10/Exploring-Neural-Networks-Quantizationvia-Layer-Wise-Quantization-Analysis.pdf - Resource Types: Article Recovering Neural Network Quantization Error Through Weight Factorization --- - Published: 2021-03-21 - Modified: 2024-11-06 - URL: https://hailo.ai/wp-content/uploads/2023/10/Exploring-Neural-Networks-Quantizationvia-Layer-Wise-Quantization-Analysis.pdf - Resource Types: Article Recovering Neural Network Quantization Error Through Weight Factorization --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2020-12-23 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=2vIm6tIhvNw - Resource Types: Presentations EVS 2020: Lessons Learned from Deep Learning Application Deployments In Edge Devices --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2020-12-23 - Modified: 2024-11-20 - URL: https://www.youtube.com/watch?v=2vIm6tIhvNw - Resource Types: Presentation EVS 2020: Lessons Learned from Deep Learning Application Deployments In Edge Devices --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2020-12-23 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=2vIm6tIhvNw - Resource Types: Presentations EVS 2020: Lessons Learned from Deep Learning Application Deployments In Edge Devices --- > Hailo's Orr Danon Shares Lessons Learned from Deep Learning Application Deployments In Edge Devices - Published: 2020-12-23 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=2vIm6tIhvNw - Resource Types: Presentations EVS 2020: Lessons Learned from Deep Learning Application Deployments In Edge Devices --- - Published: 2020-09-07 - Modified: 2024-11-06 - URL: https://youtu.be/cbHoR9_e40g - Resource Types: Presentations Liran Bar, VP of Business Development at Hailo talks about the way Hailo faces global challenges with revolutionary AI Technology. --- - Published: 2020-09-07 - Modified: 2024-11-20 - URL: https://youtu.be/cbHoR9_e40g - Resource Types: Presentation Liran Bar, VP of Business Development at Hailo talks about the way Hailo faces global challenges with revolutionary AI Technology. --- - Published: 2020-09-07 - Modified: 2024-11-06 - URL: https://youtu.be/cbHoR9_e40g - Resource Types: Presentations Liran Bar, VP of Business Development at Hailo talks about the way Hailo faces global challenges with revolutionary AI Technology. --- - Published: 2020-09-07 - Modified: 2024-11-06 - URL: https://youtu.be/cbHoR9_e40g - Resource Types: Presentations Liran Bar, VP of Business Development at Hailo talks about the way Hailo faces global challenges with revolutionary AI Technology. --- - Published: 2020-07-15 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=XsL_U0s5OMY&__hstc=&__hssc=&hsCtaTracking=d15d86fd-66c3-4ebc-a38f-bfa5f47537b7%7Ca4aec7e7-f71f-43b1-8582-744025a72e1c - Resource Types: Presentations ML Meetup 2019: What Makes Neural Networks A Unique Computational Problem? --- - Published: 2020-07-15 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=TwVVC5wVlVk&__hstc=&__hssc=&hsCtaTracking=fc4001f6-091f-49a7-bea0-9fe24b13e219%7Cc03ee7e4-6850-476a-a9cc-575bae348159 - Resource Types: Presentations ML Meetup 2019: 8-Bit Deep Learning, New Methods for Improving Quantization --- - Published: 2020-07-15 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=OCIxzgQ3fwQ&t=12s&__hstc=&__hssc=&hsCtaTracking=dd982951-33c9-4ccc-8d77-853be9302b7c%7Cf0a573e6-4422-4c0b-b171-ead7de5392a4 - Resource Types: Presentations Hailo’s Orr Danon Explores Architectural Concepts Underlying Diverse Processors (Preview) --- - Published: 2020-07-15 - Modified: 2024-11-05 - URL: https://www.youtube.com/watch?v=DhHyshhc1lY - Resource Types: Webinar Quantization of Neural Networks – High Accuracy at Low Precision --- - Published: 2020-07-15 - Modified: 2024-11-20 - URL: https://www.youtube.com/watch?v=DhHyshhc1lY - Resource Types: Webinar Quantization of Neural Networks – High Accuracy at Low Precision --- - Published: 2020-07-15 - Modified: 2024-11-20 - URL: https://www.youtube.com/watch?v=OCIxzgQ3fwQ&t=12s&__hstc=&__hssc=&hsCtaTracking=dd982951-33c9-4ccc-8d77-853be9302b7c%7Cf0a573e6-4422-4c0b-b171-ead7de5392a4 - Resource Types: Presentation Hailo’s Orr Danon Explores Architectural Concepts Underlying Diverse Processors (Preview) --- - Published: 2020-07-15 - Modified: 2024-11-20 - URL: https://www.youtube.com/watch?v=TwVVC5wVlVk&__hstc=&__hssc=&hsCtaTracking=fc4001f6-091f-49a7-bea0-9fe24b13e219%7Cc03ee7e4-6850-476a-a9cc-575bae348159 - Resource Types: Presentation ML Meetup 2019: 8-Bit Deep Learning, New Methods for Improving Quantization --- - Published: 2020-07-15 - Modified: 2024-11-20 - URL: https://www.youtube.com/watch?v=XsL_U0s5OMY&__hstc=&__hssc=&hsCtaTracking=d15d86fd-66c3-4ebc-a38f-bfa5f47537b7%7Ca4aec7e7-f71f-43b1-8582-744025a72e1c - Resource Types: Presentation ML Meetup 2019: What Makes Neural Networks A Unique Computational Problem? --- - Published: 2020-07-15 - Modified: 2024-11-05 - URL: https://www.youtube.com/watch?v=DhHyshhc1lY - Resource Types: Webinar Quantization of Neural Networks – High Accuracy at Low Precision --- - Published: 2020-07-15 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=OCIxzgQ3fwQ&t=12s&__hstc=&__hssc=&hsCtaTracking=dd982951-33c9-4ccc-8d77-853be9302b7c%7Cf0a573e6-4422-4c0b-b171-ead7de5392a4 - Resource Types: Presentations Hailo’s Orr Danon Explores Architectural Concepts Underlying Diverse Processors (Preview) --- - Published: 2020-07-15 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=TwVVC5wVlVk&__hstc=&__hssc=&hsCtaTracking=fc4001f6-091f-49a7-bea0-9fe24b13e219%7Cc03ee7e4-6850-476a-a9cc-575bae348159 - Resource Types: Presentations ML Meetup 2019: 8-Bit Deep Learning, New Methods for Improving Quantization --- - Published: 2020-07-15 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=XsL_U0s5OMY&__hstc=&__hssc=&hsCtaTracking=d15d86fd-66c3-4ebc-a38f-bfa5f47537b7%7Ca4aec7e7-f71f-43b1-8582-744025a72e1c - Resource Types: Presentations ML Meetup 2019: What Makes Neural Networks A Unique Computational Problem? --- - Published: 2020-07-15 - Modified: 2024-11-05 - URL: https://www.youtube.com/watch?v=DhHyshhc1lY - Resource Types: Webinar Quantization of Neural Networks – High Accuracy at Low Precision --- - Published: 2020-07-15 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=TwVVC5wVlVk&__hstc=&__hssc=&hsCtaTracking=fc4001f6-091f-49a7-bea0-9fe24b13e219%7Cc03ee7e4-6850-476a-a9cc-575bae348159 - Resource Types: Presentations ML Meetup 2019: 8-Bit Deep Learning, New Methods for Improving Quantization --- - Published: 2020-07-15 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=XsL_U0s5OMY&__hstc=&__hssc=&hsCtaTracking=d15d86fd-66c3-4ebc-a38f-bfa5f47537b7%7Ca4aec7e7-f71f-43b1-8582-744025a72e1c - Resource Types: Presentations ML Meetup 2019: What Makes Neural Networks A Unique Computational Problem? --- - Published: 2020-07-15 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=OCIxzgQ3fwQ&t=12s&__hstc=&__hssc=&hsCtaTracking=dd982951-33c9-4ccc-8d77-853be9302b7c%7Cf0a573e6-4422-4c0b-b171-ead7de5392a4 - Resource Types: Presentations Hailo’s Orr Danon Explores Architectural Concepts Underlying Diverse Processors (Preview) --- - Published: 2020-06-23 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=h0oTeQy-_8I&t=60s&__hstc=&__hssc=&hsCtaTracking=2375c979-5cd5-4f82-b6f9-c3f1d51c272b%7C66f82fb4-a472-46c4-be87-a860aefa9a70 - Resource Types: Presentations MWC Shanghai 2019: The Big Think AI Debate --- - Published: 2020-06-23 - Modified: 2024-11-20 - URL: https://www.youtube.com/watch?v=h0oTeQy-_8I&t=60s&__hstc=&__hssc=&hsCtaTracking=2375c979-5cd5-4f82-b6f9-c3f1d51c272b%7C66f82fb4-a472-46c4-be87-a860aefa9a70 - Resource Types: Presentation MWC Shanghai 2019: The Big Think AI Debate --- - Published: 2020-06-23 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=h0oTeQy-_8I&t=60s&__hstc=&__hssc=&hsCtaTracking=2375c979-5cd5-4f82-b6f9-c3f1d51c272b%7C66f82fb4-a472-46c4-be87-a860aefa9a70 - Resource Types: Presentations MWC Shanghai 2019: The Big Think AI Debate --- - Published: 2020-06-23 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=h0oTeQy-_8I&t=60s&__hstc=&__hssc=&hsCtaTracking=2375c979-5cd5-4f82-b6f9-c3f1d51c272b%7C66f82fb4-a472-46c4-be87-a860aefa9a70 - Resource Types: Presentations MWC Shanghai 2019: The Big Think AI Debate --- - Published: 2020-06-22 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=RjChtVvTCvQ&list=PLkNLzI6eQphI3ynBM0jnnbyCUtNfUyaUe&__hstc=&__hssc=&hsCtaTracking=ec6b012f-9df7-43a9-9ca6-c2543a807e03%7Ccd0fd2b0-9082-4ff6-8135-361373cf751b - Resource Types: Presentations MWC Shanghai 2019: What can AI chipmakers do? --- - Published: 2020-06-22 - Modified: 2024-11-20 - URL: https://www.youtube.com/watch?v=RjChtVvTCvQ&list=PLkNLzI6eQphI3ynBM0jnnbyCUtNfUyaUe&__hstc=&__hssc=&hsCtaTracking=ec6b012f-9df7-43a9-9ca6-c2543a807e03%7Ccd0fd2b0-9082-4ff6-8135-361373cf751b - Resource Types: Presentation MWC Shanghai 2019: What can AI chipmakers do? --- - Published: 2020-06-22 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=RjChtVvTCvQ&list=PLkNLzI6eQphI3ynBM0jnnbyCUtNfUyaUe&__hstc=&__hssc=&hsCtaTracking=ec6b012f-9df7-43a9-9ca6-c2543a807e03%7Ccd0fd2b0-9082-4ff6-8135-361373cf751b - Resource Types: Presentations MWC Shanghai 2019: What can AI chipmakers do? --- - Published: 2020-06-22 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=RjChtVvTCvQ&list=PLkNLzI6eQphI3ynBM0jnnbyCUtNfUyaUe&__hstc=&__hssc=&hsCtaTracking=ec6b012f-9df7-43a9-9ca6-c2543a807e03%7Ccd0fd2b0-9082-4ff6-8135-361373cf751b - Resource Types: Presentations MWC Shanghai 2019: What can AI chipmakers do? --- - Published: 2020-06-18 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=oZ1_nrXcbeM&__hstc=&__hssc=&hsCtaTracking=26310300-682e-4959-994a-50e22851b2cf%7C00c405f5-b0f4-4064-bd9c-8c8b866ae035 - Resource Types: Presentations AutoSens Brussels 2018: State-of-the-art processor by Hailo --- - Published: 2020-06-18 - Modified: 2024-11-20 - URL: https://www.youtube.com/watch?v=oZ1_nrXcbeM&__hstc=&__hssc=&hsCtaTracking=26310300-682e-4959-994a-50e22851b2cf%7C00c405f5-b0f4-4064-bd9c-8c8b866ae035 - Resource Types: Presentation AutoSens Brussels 2018: State-of-the-art processor by Hailo --- - Published: 2020-06-18 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=oZ1_nrXcbeM&__hstc=&__hssc=&hsCtaTracking=26310300-682e-4959-994a-50e22851b2cf%7C00c405f5-b0f4-4064-bd9c-8c8b866ae035 - Resource Types: Presentations AutoSens Brussels 2018: State-of-the-art processor by Hailo --- - Published: 2020-06-18 - Modified: 2024-11-06 - URL: https://www.youtube.com/watch?v=oZ1_nrXcbeM&__hstc=&__hssc=&hsCtaTracking=26310300-682e-4959-994a-50e22851b2cf%7C00c405f5-b0f4-4064-bd9c-8c8b866ae035 - Resource Types: Presentations AutoSens Brussels 2018: State-of-the-art processor by Hailo --- --- ## Products ---