ImageNet

Network Name Accuracy (top1)QuantizedInput Resolution (HxWxC)Params (M)FLOPs (G)PretrainedSourceCompiled
efficientnet_l 80.4679.36300x300x310.559.70link link link
efficientnet_lite0 74.9973.91224x224x34.630.39link link link
efficientnet_lite1 76.6876.36240x240x35.390.61link link link
efficientnet_lite2 77.4576.74260x260x36.060.87link link link
efficientnet_lite3 79.2978.71280x280x38.161.40link link link
efficientnet_lite4 80.7980.47300x300x312.952.58link link link
efficientnet_m 🚀 78.9178.63240x240x36.873.68link link link
efficientnet_s 77.6477.32224x224x35.412.36link link link
hardnet39ds 73.4372.23224x224x33.480.43link link link
hardnet68 75.4775.04224x224x317.564.25link link link
inception_v1 69.7469.3224x224x36.621.50link link link
mobilenet_v1 70.9770.25224x224x34.220.57link link link
mobilenet_v2_1.0 🚀 71.7870.64224x224x33.490.31link link link
mobilenet_v2_1.4 74.1873.07224x224x36.090.59link link link
mobilenet_v3 72.2171.73224x224x34.071.00link link link
mobilenet_v3_large_minimalistic 🚀 72.1171.24224x224x33.910.21link link link
regnetx_1.6gf 77.0576.44224x224x39.171.61link link link
regnetx_800mf 🚀 75.1674.51224x224x37.240.80link link link
regnety_200mf 70.3869.91224x224x33.150.20link link link
resmlp12_relu 75.2674.06224x224x315.773.02link link link
resnet_v1_18 71.2670.64224x224x311.681.82link link link
resnet_v1_34 72.772.14224x224x321.793.67link link link
resnet_v1_50 🚀 ⭐ 75.1274.47224x224x325.533.49link link link
resnet_v2_18 69.5769.1224x224x311.681.82link link link
resnet_v2_34 73.0772.72224x224x321.793.67link link link
resnext26_32x4d 76.1875.78224x224x315.372.48link link link
resnext50_32x4d 79.3178.39224x224x324.994.24link link link
shufflenet_g8_w1 66.365.44224x224x32.460.18link link link
squeezenet_v1.1 🚀 59.8559.4224x224x31.240.39link link link
  • Network available in Hailo Benchmark are marked with 🚀
  • Networks available in TAPPAS are marked with ⭐