EfficientNet_and_ResNeXt101_wsl.md 9.7 KB
Newer Older
W
WuHaobo 已提交
1 2 3
# EfficientNet与ResNeXt101_wsl系列

## 概述
littletomatodonkey's avatar
littletomatodonkey 已提交
4
EfficientNet是Google于2019年发布的一个基于NAS的轻量级网络,其中EfficientNetB7刷新了当时ImageNet-1k的分类准确率。在该文章中,作者指出,传统的提升神经网络性能的方法主要是从网络的宽度、网络的深度、以及输入图片的分辨率入手,但是作者通过实验发现,平衡这三个维度对精度和效率的提升至关重要,于是,作者通过一系列的实验中总结出了如何同时平衡这三个维度的放缩,与此同时,基于这种放缩方法,作者在EfficientNet_B0的基础上,构建了EfficientNet系列中B1-B7共7个网络,并在同样FLOPS与参数量的情况下,精度达到了state-of-the-art的效果。
littletomatodonkey's avatar
littletomatodonkey 已提交
5

littletomatodonkey's avatar
littletomatodonkey 已提交
6
ResNeXt是facebook于2016年提出的一种对ResNet的改进版网络。在2019年,facebook通过弱监督学习研究了该系列网络在ImageNet上的精度上限,为了区别之前的ResNeXt网络,该系列网络的后缀为wsl,其中wsl是弱监督学习(weakly-supervised-learning)的简称。为了能有更强的特征提取能力,研究者将其网络宽度进一步放大,其中最大的ResNeXt101_32x48d_wsl拥有8亿个参数,将其在9.4亿的弱标签图片下训练并在ImageNet-1k上做finetune,最终在ImageNet-1k的top-1达到了85.4%,这也是迄今为止在ImageNet-1k的数据集上以224x224的分辨率下精度最高的网络。Fix-ResNeXt中,作者使用了更大的图像分辨率,针对训练图片和验证图片数据预处理不一致的情况下做了专门的Fix策略,并使得ResNeXt101_32x48d_wsl拥有了更高的精度,由于其用到了Fix策略,故命名为Fix-ResNeXt101_32x48d_wsl。
littletomatodonkey's avatar
littletomatodonkey 已提交
7

littletomatodonkey's avatar
littletomatodonkey 已提交
8

littletomatodonkey's avatar
littletomatodonkey 已提交
9
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。
littletomatodonkey's avatar
littletomatodonkey 已提交
10

littletomatodonkey's avatar
littletomatodonkey 已提交
11
![](../../images/models/T4_benchmark/t4.fp32.bs4.EfficientNet.flops.png)
littletomatodonkey's avatar
littletomatodonkey 已提交
12

littletomatodonkey's avatar
littletomatodonkey 已提交
13
![](../../images/models/T4_benchmark/t4.fp32.bs4.EfficientNet.params.png)
littletomatodonkey's avatar
littletomatodonkey 已提交
14

littletomatodonkey's avatar
littletomatodonkey 已提交
15
![](../../images/models/T4_benchmark/t4.fp32.bs1.EfficientNet.png)
W
WuHaobo 已提交
16

littletomatodonkey's avatar
littletomatodonkey 已提交
17 18
![](../../images/models/T4_benchmark/t4.fp16.bs1.EfficientNet.png)

littletomatodonkey's avatar
littletomatodonkey 已提交
19
目前PaddleClas开源的这两类模型的预训练模型一共有14个。从上图中可以看出EfficientNet系列网络优势非常明显,ResNeXt101_wsl系列模型由于用到了更多的数据,最终的精度也更高。EfficientNet_B0_Small是去掉了SE_block的EfficientNet_B0,其具有更快的推理速度。
W
WuHaobo 已提交
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

## 精度、FLOPS和参数量

| Models                        | Top1   | Top5   | Reference<br>top1 | Reference<br>top5 | FLOPS<br>(G) | Parameters<br>(M) |
|:--:|:--:|:--:|:--:|:--:|:--:|:--:|
| ResNeXt101_<br>32x8d_wsl      | 0.826  | 0.967  | 0.822             | 0.964             | 29.140       | 78.440            |
| ResNeXt101_<br>32x16d_wsl     | 0.842  | 0.973  | 0.842             | 0.972             | 57.550       | 152.660           |
| ResNeXt101_<br>32x32d_wsl     | 0.850  | 0.976  | 0.851             | 0.975             | 115.170      | 303.110           |
| ResNeXt101_<br>32x48d_wsl     | 0.854  | 0.977  | 0.854             | 0.976             | 173.580      | 456.200           |
| Fix_ResNeXt101_<br>32x48d_wsl | 0.863  | 0.980  | 0.864             | 0.980             | 354.230      | 456.200           |
| EfficientNetB0                | 0.774  | 0.933  | 0.773             | 0.935             | 0.720        | 5.100             |
| EfficientNetB1                | 0.792  | 0.944  | 0.792             | 0.945             | 1.270        | 7.520             |
| EfficientNetB2                | 0.799  | 0.947  | 0.803             | 0.950             | 1.850        | 8.810             |
| EfficientNetB3                | 0.812  | 0.954  | 0.817             | 0.956             | 3.430        | 11.840            |
| EfficientNetB4                | 0.829  | 0.962  | 0.830             | 0.963             | 8.290        | 18.760            |
| EfficientNetB5                | 0.836  | 0.967  | 0.837             | 0.967             | 19.510       | 29.610            |
| EfficientNetB6                | 0.840  | 0.969  | 0.842             | 0.968             | 36.270       | 42.000            |
| EfficientNetB7                | 0.843  | 0.969  | 0.844             | 0.971             | 72.350       | 64.920            |
| EfficientNetB0_<br>small      | 0.758  | 0.926  |                   |                   | 0.720        | 4.650             |


littletomatodonkey's avatar
littletomatodonkey 已提交
41
## 基于V100 GPU的预测速度
W
WuHaobo 已提交
42

littletomatodonkey's avatar
littletomatodonkey 已提交
43
| Models                               | Crop Size | Resize Short Size | FP32<br>Batch Size=1<br>(ms) |
littletomatodonkey's avatar
littletomatodonkey 已提交
44 45 46 47 48 49 50 51 52 53 54 55 56 57 58
|-------------------------------|-----------|-------------------|--------------------------|
| ResNeXt101_<br>32x8d_wsl      | 224       | 256               | 19.127                   |
| ResNeXt101_<br>32x16d_wsl     | 224       | 256               | 23.629                   |
| ResNeXt101_<br>32x32d_wsl     | 224       | 256               | 40.214                   |
| ResNeXt101_<br>32x48d_wsl     | 224       | 256               | 59.714                   |
| Fix_ResNeXt101_<br>32x48d_wsl | 320       | 320               | 82.431                   |
| EfficientNetB0                | 224       | 256               | 2.449                    |
| EfficientNetB1                | 240       | 272               | 3.547                    |
| EfficientNetB2                | 260       | 292               | 3.908                    |
| EfficientNetB3                | 300       | 332               | 5.145                    |
| EfficientNetB4                | 380       | 412               | 7.609                    |
| EfficientNetB5                | 456       | 488               | 12.078                   |
| EfficientNetB6                | 528       | 560               | 18.381                   |
| EfficientNetB7                | 600       | 632               | 27.817                   |
| EfficientNetB0_<br>small      | 224       | 256               | 1.692                    |
littletomatodonkey's avatar
littletomatodonkey 已提交
59 60 61 62 63



## 基于T4 GPU的预测速度

littletomatodonkey's avatar
fix bs  
littletomatodonkey 已提交
64
| Models                    | Crop Size | Resize Short Size | FP16<br>Batch Size=1<br>(ms) | FP16<br>Batch Size=4<br>(ms) | FP16<br>Batch Size=8<br>(ms) | FP32<br>Batch Size=1<br>(ms) | FP32<br>Batch Size=4<br>(ms) | FP32<br>Batch Size=8<br>(ms) |
littletomatodonkey's avatar
littletomatodonkey 已提交
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
|---------------------------|-----------|-------------------|------------------------------|------------------------------|------------------------------|------------------------------|------------------------------|------------------------------|
| ResNeXt101_<br>32x8d_wsl      | 224       | 256               | 18.19374                     | 21.93529                     | 34.67802                     | 18.52528                     | 34.25319                     | 67.2283                      |
| ResNeXt101_<br>32x16d_wsl     | 224       | 256               | 18.52609                     | 36.8288                      | 62.79947                     | 25.60395                     | 71.88384                     | 137.62327                    |
| ResNeXt101_<br>32x32d_wsl     | 224       | 256               | 33.51391                     | 70.09682                     | 125.81884                    | 54.87396                     | 160.04337                    | 316.17718                    |
| ResNeXt101_<br>32x48d_wsl     | 224       | 256               | 50.97681                     | 137.60926                    | 190.82628                    | 99.01698256                  | 315.91261                    | 551.83695                    |
| Fix_ResNeXt101_<br>32x48d_wsl | 320       | 320               | 78.62869                     | 191.76039                    | 317.15436                    | 160.0838242                  | 595.99296                    | 1151.47384                   |
| EfficientNetB0            | 224       | 256               | 3.40122                      | 5.95851                      | 9.10801                      | 3.442                        | 6.11476                      | 9.3304                       |
| EfficientNetB1            | 240       | 272               | 5.25172                      | 9.10233                      | 14.11319                     | 5.3322                       | 9.41795                      | 14.60388                     |
| EfficientNetB2            | 260       | 292               | 5.91052                      | 10.5898                      | 17.38106                     | 6.29351                      | 10.95702                     | 17.75308                     |
| EfficientNetB3            | 300       | 332               | 7.69582                      | 16.02548                     | 27.4447                      | 7.67749                      | 16.53288                     | 28.5939                      |
| EfficientNetB4            | 380       | 412               | 11.55585                     | 29.44261                     | 53.97363                     | 12.15894                     | 30.94567                     | 57.38511                     |
| EfficientNetB5            | 456       | 488               | 19.63083                     | 56.52299                     | -                            | 20.48571                     | 61.60252                     | -                            |
| EfficientNetB6            | 528       | 560               | 30.05911                     | -                            | -                            | 32.62402                     | -                            | -                            |
| EfficientNetB7            | 600       | 632               | 47.86087                     | -                            | -                            | 53.93823                     | -                            | -                            |
| EfficientNetB0_small      | 224       | 256               | 2.39166                      | 4.36748                      | 6.96002                      | 2.3076                       | 4.71886                      | 7.21888                      |