# ResNet及其Vd系列 ## 概述 ResNet系列模型是在2015年提出的，一举在ILSVRC2015比赛中取得冠军，top5错误率为3.57%。该网络创新性的提出了残差结构，通过堆叠多个残差结构从而构建了ResNet网络。实验表明使用残差块可以有效地提升收敛速度和精度。 斯坦福大学的Joyce Xu将ResNet称为「真正重新定义了我们看待神经网络的方式」的三大架构之一。由于ResNet卓越的性能，越来越多的来自学术界和工业界学者和工程师对其结构进行了改进，比较出名的有Wide-ResNet, ResNet-vc ,ResNet-vd, Res2Net等，其中ResNet-vc与ResNet-vd的参数量和计算量与ResNet几乎一致，所以在此我们将其与ResNet统一归为ResNet系列。 本次发布ResNet系列的模型包括ResNet50，ResNet50_vd，ResNet50_vd_ssld，ResNet200_vd等14个预训练模型。在训练层面上，ResNet的模型采用了训练ImageNet的标准训练流程，而其余改进版模型采用了更多的训练策略，如learning rate的下降方式采用了cosine decay，引入了label smoothing的标签正则方式，在数据预处理加入了mixup的操作，迭代总轮数从120个epoch增加到200个epoch。 其中，ResNet50_vd_v2与ResNet50_vd_ssld采用了知识蒸馏，保证模型结构不变的情况下，进一步提升了模型的精度，具体地，ResNet50_vd_v2的teacher模型是ResNet152_vd（top1准确率80.59%），数据选用的是ImageNet-1k的训练集，ResNet50_vd_ssld的teacher模型是ResNeXt101_32x16d_wsl（top1准确率84.2%），数据选用结合了ImageNet-1k的训练集和ImageNet-22k挖掘的400万数据。知识蒸馏的具体方法正在持续更新中。 该系列模型的FLOPS、参数量以及T4 GPU上的预测耗时如下图所示。 ![](../../images/models/T4_benchmark/t4.fp32.bs4.ResNet.flops.png) ![](../../images/models/T4_benchmark/t4.fp32.bs4.ResNet.params.png) ![](../../images/models/T4_benchmark/t4.fp32.bs4.ResNet.png) ![](../../images/models/T4_benchmark/t4.fp16.bs4.ResNet.png) 通过上述曲线可以看出，层数越多，准确率越高，但是相应的参数量、计算量和延时都会增加。ResNet50_vd_ssld通过用更强的teacher和更多的数据，将其在ImageNet-1k上的验证集top-1精度进一步提高，达到了82.39%，刷新了ResNet50系列模型的精度。 ## 精度、FLOPS和参数量 | Models | Top1 | Top5 | Reference
top1 | Reference
top5 | FLOPS
(G) | Parameters
(M) | |:--:|:--:|:--:|:--:|:--:|:--:|:--:| | ResNet18 | 0.710 | 0.899 | 0.696 | 0.891 | 3.660 | 11.690 | | ResNet18_vd | 0.723 | 0.908 | | | 4.140 | 11.710 | | ResNet34 | 0.746 | 0.921 | 0.732 | 0.913 | 7.360 | 21.800 | | ResNet34_vd | 0.760 | 0.930 | | | 7.390 | 21.820 | | ResNet34_vd_ssld | 0.797 | 0.949 | | | 7.390 | 21.820 | | ResNet50 | 0.765 | 0.930 | 0.760 | 0.930 | 8.190 | 25.560 | | ResNet50_vc | 0.784 | 0.940 | | | 8.670 | 25.580 | | ResNet50_vd | 0.791 | 0.944 | 0.792 | 0.946 | 8.670 | 25.580 | | ResNet50_vd_v2 | 0.798 | 0.949 | | | 8.670 | 25.580 | | ResNet101 | 0.776 | 0.936 | 0.776 | 0.938 | 15.520 | 44.550 | | ResNet101_vd | 0.802 | 0.950 | | | 16.100 | 44.570 | | ResNet152 | 0.783 | 0.940 | 0.778 | 0.938 | 23.050 | 60.190 | | ResNet152_vd | 0.806 | 0.953 | | | 23.530 | 60.210 | | ResNet200_vd | 0.809 | 0.953 | | | 30.530 | 74.740 | | ResNet50_vd_ssld | 0.824 | 0.961 | | | 8.670 | 25.580 | | ResNet50_vd_ssld_v2 | 0.830 | 0.964 | | | 8.670 | 25.580 | | Fix_ResNet50_vd_ssld_v2 | 0.840 | 0.970 | | | 17.696 | 25.580 | | ResNet101_vd_ssld | 0.837 | 0.967 | | | 16.100 | 44.570 | * 注：`ResNet50_vd_ssld_v2`是在`ResNet50_vd_ssld`训练策略的基础上加上AutoAugment训练得到，`Fix_ResNet50_vd_ssld_v2`是固定`ResNet50_vd_ssld_v2`除FC层外所有的网络参数，在320x320的图像输入分辨率下，基于ImageNet1k数据集微调得到。 ## 基于V100 GPU的预测速度 | Models | Crop Size | Resize Short Size | FP32
Batch Size=1
(ms) | |------------------|-----------|-------------------|--------------------------| | ResNet18 | 224 | 256 | 1.499 | | ResNet18_vd | 224 | 256 | 1.603 | | ResNet34 | 224 | 256 | 2.272 | | ResNet34_vd | 224 | 256 | 2.343 | | ResNet34_vd_ssld | 224 | 256 | 2.343 | | ResNet50 | 224 | 256 | 2.939 | | ResNet50_vc | 224 | 256 | 3.041 | | ResNet50_vd | 224 | 256 | 3.165 | | ResNet50_vd_v2 | 224 | 256 | 3.165 | | ResNet101 | 224 | 256 | 5.314 | | ResNet101_vd | 224 | 256 | 5.252 | | ResNet152 | 224 | 256 | 7.205 | | ResNet152_vd | 224 | 256 | 7.200 | | ResNet200_vd | 224 | 256 | 8.885 | | ResNet50_vd_ssld | 224 | 256 | 3.165 | | ResNet101_vd_ssld | 224 | 256 | 5.252 | ## 基于T4 GPU的预测速度 | Models | Crop Size | Resize Short Size | FP16
Batch Size=1
(ms) | FP16
Batch Size=4
(ms) | FP16
Batch Size=8
(ms) | FP32
Batch Size=1
(ms) | FP32
Batch Size=4
(ms) | FP32
Batch Size=8
(ms) | |-------------------|-----------|-------------------|------------------------------|------------------------------|------------------------------|------------------------------|------------------------------|------------------------------| | ResNet18 | 224 | 256 | 1.3568 | 2.5225 | 3.61904 | 1.45606 | 3.56305 | 6.28798 | | ResNet18_vd | 224 | 256 | 1.39593 | 2.69063 | 3.88267 | 1.54557 | 3.85363 | 6.88121 | | ResNet34 | 224 | 256 | 2.23092 | 4.10205 | 5.54904 | 2.34957 | 5.89821 | 10.73451 | | ResNet34_vd | 224 | 256 | 2.23992 | 4.22246 | 5.79534 | 2.43427 | 6.22257 | 11.44906 | | ResNet34_vd_ssld | 224 | 256 | 2.23992 | 4.22246 | 5.79534 | 2.43427 | 6.22257 | 11.44906 | | ResNet50 | 224 | 256 | 2.63824 | 4.63802 | 7.02444 | 3.47712 | 7.84421 | 13.90633 | | ResNet50_vc | 224 | 256 | 2.67064 | 4.72372 | 7.17204 | 3.52346 | 8.10725 | 14.45577 | | ResNet50_vd | 224 | 256 | 2.65164 | 4.84109 | 7.46225 | 3.53131 | 8.09057 | 14.45965 | | ResNet50_vd_v2 | 224 | 256 | 2.65164 | 4.84109 | 7.46225 | 3.53131 | 8.09057 | 14.45965 | | ResNet101 | 224 | 256 | 5.04037 | 7.73673 | 10.8936 | 6.07125 | 13.40573 | 24.3597 | | ResNet101_vd | 224 | 256 | 5.05972 | 7.83685 | 11.34235 | 6.11704 | 13.76222 | 25.11071 | | ResNet152 | 224 | 256 | 7.28665 | 10.62001 | 14.90317 | 8.50198 | 19.17073 | 35.78384 | | ResNet152_vd | 224 | 256 | 7.29127 | 10.86137 | 15.32444 | 8.54376 | 19.52157 | 36.64445 | | ResNet200_vd | 224 | 256 | 9.36026 | 13.5474 | 19.0725 | 10.80619 | 25.01731 | 48.81399 | | ResNet50_vd_ssld | 224 | 256 | 2.65164 | 4.84109 | 7.46225 | 3.53131 | 8.09057 | 14.45965 | | ResNet50_vd_ssld_v2 | 224 | 256 | 2.65164 | 4.84109 | 7.46225 | 3.53131 | 8.09057 | 14.45965 | | Fix_ResNet50_vd_ssld_v2 | 320 | 320 | 3.42818 | 7.51534 | 13.19370 | 5.07696 | 14.64218 | 27.01453 | | ResNet101_vd_ssld | 224 | 256 | 5.05972 | 7.83685 | 11.34235 | 6.11704 | 13.76222 | 25.11071 |