You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
385 lines
9.4 KiB
385 lines
9.4 KiB
3 years ago
|
# AdvProp
|
||
3 years ago
|
|
||
|
**AdvProp** is an adversarial training scheme which treats adversarial examples as additional examples, to prevent overfitting. Key to the method is the usage of a separate auxiliary batch norm for adversarial examples, as they have different underlying distributions to normal examples.
|
||
|
|
||
|
{% include 'code_snippets.md' %}
|
||
|
|
||
|
## How do I train this model?
|
||
|
|
||
|
You can follow the [timm recipe scripts](https://rwightman.github.io/pytorch-image-models/scripts/) for training a new model afresh.
|
||
|
|
||
|
## Citation
|
||
|
|
||
|
```BibTeX
|
||
|
@misc{xie2020adversarial,
|
||
|
title={Adversarial Examples Improve Image Recognition},
|
||
|
author={Cihang Xie and Mingxing Tan and Boqing Gong and Jiang Wang and Alan Yuille and Quoc V. Le},
|
||
|
year={2020},
|
||
|
eprint={1911.09665},
|
||
|
archivePrefix={arXiv},
|
||
|
primaryClass={cs.CV}
|
||
|
}
|
||
|
```
|
||
|
|
||
|
<!--
|
||
|
Models:
|
||
|
- Name: tf_efficientnet_b1_ap
|
||
|
Metadata:
|
||
|
FLOPs: 883633200
|
||
|
Epochs: 350
|
||
|
Batch Size: 2048
|
||
|
Training Data:
|
||
|
- ImageNet
|
||
|
Training Techniques:
|
||
|
- AdvProp
|
||
|
- AutoAugment
|
||
|
- Label Smoothing
|
||
|
- RMSProp
|
||
|
- Stochastic Depth
|
||
|
- Weight Decay
|
||
|
Architecture:
|
||
|
- 1x1 Convolution
|
||
|
- Average Pooling
|
||
|
- Batch Normalization
|
||
|
- Convolution
|
||
|
- Dense Connections
|
||
|
- Dropout
|
||
|
- Inverted Residual Block
|
||
|
- Squeeze-and-Excitation Block
|
||
|
- Swish
|
||
|
File Size: 31515350
|
||
|
Tasks:
|
||
|
- Image Classification
|
||
|
ID: tf_efficientnet_b1_ap
|
||
|
LR: 0.256
|
||
|
Crop Pct: '0.882'
|
||
|
Momentum: 0.9
|
||
|
Image Size: '240'
|
||
|
Weight Decay: 1.0e-05
|
||
|
Interpolation: bicubic
|
||
|
RMSProp Decay: 0.9
|
||
|
Label Smoothing: 0.1
|
||
|
BatchNorm Momentum: 0.99
|
||
|
Code: https://github.com/rwightman/pytorch-image-models/blob/9a25fdf3ad0414b4d66da443fe60ae0aa14edc84/timm/models/efficientnet.py#L1344
|
||
|
In Collection: AdvProp
|
||
|
- Name: tf_efficientnet_b2_ap
|
||
|
Metadata:
|
||
|
FLOPs: 1234321170
|
||
|
Epochs: 350
|
||
|
Batch Size: 2048
|
||
|
Training Data:
|
||
|
- ImageNet
|
||
|
Training Techniques:
|
||
|
- AdvProp
|
||
|
- AutoAugment
|
||
|
- Label Smoothing
|
||
|
- RMSProp
|
||
|
- Stochastic Depth
|
||
|
- Weight Decay
|
||
|
Architecture:
|
||
|
- 1x1 Convolution
|
||
|
- Average Pooling
|
||
|
- Batch Normalization
|
||
|
- Convolution
|
||
|
- Dense Connections
|
||
|
- Dropout
|
||
|
- Inverted Residual Block
|
||
|
- Squeeze-and-Excitation Block
|
||
|
- Swish
|
||
|
File Size: 36800745
|
||
|
Tasks:
|
||
|
- Image Classification
|
||
|
ID: tf_efficientnet_b2_ap
|
||
|
LR: 0.256
|
||
|
Crop Pct: '0.89'
|
||
|
Momentum: 0.9
|
||
|
Image Size: '260'
|
||
|
Weight Decay: 1.0e-05
|
||
|
Interpolation: bicubic
|
||
|
RMSProp Decay: 0.9
|
||
|
Label Smoothing: 0.1
|
||
|
BatchNorm Momentum: 0.99
|
||
|
Code: https://github.com/rwightman/pytorch-image-models/blob/9a25fdf3ad0414b4d66da443fe60ae0aa14edc84/timm/models/efficientnet.py#L1354
|
||
|
In Collection: AdvProp
|
||
|
- Name: tf_efficientnet_b3_ap
|
||
|
Metadata:
|
||
|
FLOPs: 2275247568
|
||
|
Epochs: 350
|
||
|
Batch Size: 2048
|
||
|
Training Data:
|
||
|
- ImageNet
|
||
|
Training Techniques:
|
||
|
- AdvProp
|
||
|
- AutoAugment
|
||
|
- Label Smoothing
|
||
|
- RMSProp
|
||
|
- Stochastic Depth
|
||
|
- Weight Decay
|
||
|
Architecture:
|
||
|
- 1x1 Convolution
|
||
|
- Average Pooling
|
||
|
- Batch Normalization
|
||
|
- Convolution
|
||
|
- Dense Connections
|
||
|
- Dropout
|
||
|
- Inverted Residual Block
|
||
|
- Squeeze-and-Excitation Block
|
||
|
- Swish
|
||
|
File Size: 49384538
|
||
|
Tasks:
|
||
|
- Image Classification
|
||
|
ID: tf_efficientnet_b3_ap
|
||
|
LR: 0.256
|
||
|
Crop Pct: '0.904'
|
||
|
Momentum: 0.9
|
||
|
Image Size: '300'
|
||
|
Weight Decay: 1.0e-05
|
||
|
Interpolation: bicubic
|
||
|
RMSProp Decay: 0.9
|
||
|
Label Smoothing: 0.1
|
||
|
BatchNorm Momentum: 0.99
|
||
|
Code: https://github.com/rwightman/pytorch-image-models/blob/9a25fdf3ad0414b4d66da443fe60ae0aa14edc84/timm/models/efficientnet.py#L1364
|
||
|
In Collection: AdvProp
|
||
|
- Name: tf_efficientnet_b4_ap
|
||
|
Metadata:
|
||
|
FLOPs: 5749638672
|
||
|
Epochs: 350
|
||
|
Batch Size: 2048
|
||
|
Training Data:
|
||
|
- ImageNet
|
||
|
Training Techniques:
|
||
|
- AdvProp
|
||
|
- AutoAugment
|
||
|
- Label Smoothing
|
||
|
- RMSProp
|
||
|
- Stochastic Depth
|
||
|
- Weight Decay
|
||
|
Architecture:
|
||
|
- 1x1 Convolution
|
||
|
- Average Pooling
|
||
|
- Batch Normalization
|
||
|
- Convolution
|
||
|
- Dense Connections
|
||
|
- Dropout
|
||
|
- Inverted Residual Block
|
||
|
- Squeeze-and-Excitation Block
|
||
|
- Swish
|
||
|
File Size: 77993585
|
||
|
Tasks:
|
||
|
- Image Classification
|
||
|
ID: tf_efficientnet_b4_ap
|
||
|
LR: 0.256
|
||
|
Crop Pct: '0.922'
|
||
|
Momentum: 0.9
|
||
|
Image Size: '380'
|
||
|
Weight Decay: 1.0e-05
|
||
|
Interpolation: bicubic
|
||
|
RMSProp Decay: 0.9
|
||
|
Label Smoothing: 0.1
|
||
|
BatchNorm Momentum: 0.99
|
||
|
Code: https://github.com/rwightman/pytorch-image-models/blob/9a25fdf3ad0414b4d66da443fe60ae0aa14edc84/timm/models/efficientnet.py#L1374
|
||
|
In Collection: AdvProp
|
||
|
- Name: tf_efficientnet_b5_ap
|
||
|
Metadata:
|
||
|
FLOPs: 13176501888
|
||
|
Epochs: 350
|
||
|
Batch Size: 2048
|
||
|
Training Data:
|
||
|
- ImageNet
|
||
|
Training Techniques:
|
||
|
- AdvProp
|
||
|
- AutoAugment
|
||
|
- Label Smoothing
|
||
|
- RMSProp
|
||
|
- Stochastic Depth
|
||
|
- Weight Decay
|
||
|
Architecture:
|
||
|
- 1x1 Convolution
|
||
|
- Average Pooling
|
||
|
- Batch Normalization
|
||
|
- Convolution
|
||
|
- Dense Connections
|
||
|
- Dropout
|
||
|
- Inverted Residual Block
|
||
|
- Squeeze-and-Excitation Block
|
||
|
- Swish
|
||
|
File Size: 122403150
|
||
|
Tasks:
|
||
|
- Image Classification
|
||
|
ID: tf_efficientnet_b5_ap
|
||
|
LR: 0.256
|
||
|
Crop Pct: '0.934'
|
||
|
Momentum: 0.9
|
||
|
Image Size: '456'
|
||
|
Weight Decay: 1.0e-05
|
||
|
Interpolation: bicubic
|
||
|
RMSProp Decay: 0.9
|
||
|
Label Smoothing: 0.1
|
||
|
BatchNorm Momentum: 0.99
|
||
|
Code: https://github.com/rwightman/pytorch-image-models/blob/9a25fdf3ad0414b4d66da443fe60ae0aa14edc84/timm/models/efficientnet.py#L1384
|
||
|
In Collection: AdvProp
|
||
|
- Name: tf_efficientnet_b6_ap
|
||
|
Metadata:
|
||
|
FLOPs: 24180518488
|
||
|
Epochs: 350
|
||
|
Batch Size: 2048
|
||
|
Training Data:
|
||
|
- ImageNet
|
||
|
Training Techniques:
|
||
|
- AdvProp
|
||
|
- AutoAugment
|
||
|
- Label Smoothing
|
||
|
- RMSProp
|
||
|
- Stochastic Depth
|
||
|
- Weight Decay
|
||
|
Architecture:
|
||
|
- 1x1 Convolution
|
||
|
- Average Pooling
|
||
|
- Batch Normalization
|
||
|
- Convolution
|
||
|
- Dense Connections
|
||
|
- Dropout
|
||
|
- Inverted Residual Block
|
||
|
- Squeeze-and-Excitation Block
|
||
|
- Swish
|
||
|
File Size: 173237466
|
||
|
Tasks:
|
||
|
- Image Classification
|
||
|
ID: tf_efficientnet_b6_ap
|
||
|
LR: 0.256
|
||
|
Crop Pct: '0.942'
|
||
|
Momentum: 0.9
|
||
|
Image Size: '528'
|
||
|
Weight Decay: 1.0e-05
|
||
|
Interpolation: bicubic
|
||
|
RMSProp Decay: 0.9
|
||
|
Label Smoothing: 0.1
|
||
|
BatchNorm Momentum: 0.99
|
||
|
Code: https://github.com/rwightman/pytorch-image-models/blob/9a25fdf3ad0414b4d66da443fe60ae0aa14edc84/timm/models/efficientnet.py#L1394
|
||
|
In Collection: AdvProp
|
||
|
- Name: tf_efficientnet_b7_ap
|
||
|
Metadata:
|
||
|
FLOPs: 48205304880
|
||
|
Epochs: 350
|
||
|
Batch Size: 2048
|
||
|
Training Data:
|
||
|
- ImageNet
|
||
|
Training Techniques:
|
||
|
- AdvProp
|
||
|
- AutoAugment
|
||
|
- Label Smoothing
|
||
|
- RMSProp
|
||
|
- Stochastic Depth
|
||
|
- Weight Decay
|
||
|
Architecture:
|
||
|
- 1x1 Convolution
|
||
|
- Average Pooling
|
||
|
- Batch Normalization
|
||
|
- Convolution
|
||
|
- Dense Connections
|
||
|
- Dropout
|
||
|
- Inverted Residual Block
|
||
|
- Squeeze-and-Excitation Block
|
||
|
- Swish
|
||
|
File Size: 266850607
|
||
|
Tasks:
|
||
|
- Image Classification
|
||
|
ID: tf_efficientnet_b7_ap
|
||
|
LR: 0.256
|
||
|
Crop Pct: '0.949'
|
||
|
Momentum: 0.9
|
||
|
Image Size: '600'
|
||
|
Weight Decay: 1.0e-05
|
||
|
Interpolation: bicubic
|
||
|
RMSProp Decay: 0.9
|
||
|
Label Smoothing: 0.1
|
||
|
BatchNorm Momentum: 0.99
|
||
|
Code: https://github.com/rwightman/pytorch-image-models/blob/9a25fdf3ad0414b4d66da443fe60ae0aa14edc84/timm/models/efficientnet.py#L1405
|
||
|
In Collection: AdvProp
|
||
|
- Name: tf_efficientnet_b8_ap
|
||
|
Metadata:
|
||
|
FLOPs: 80962956270
|
||
|
Epochs: 350
|
||
|
Batch Size: 2048
|
||
|
Training Data:
|
||
|
- ImageNet
|
||
|
Training Techniques:
|
||
|
- AdvProp
|
||
|
- AutoAugment
|
||
|
- Label Smoothing
|
||
|
- RMSProp
|
||
|
- Stochastic Depth
|
||
|
- Weight Decay
|
||
|
Architecture:
|
||
|
- 1x1 Convolution
|
||
|
- Average Pooling
|
||
|
- Batch Normalization
|
||
|
- Convolution
|
||
|
- Dense Connections
|
||
|
- Dropout
|
||
|
- Inverted Residual Block
|
||
|
- Squeeze-and-Excitation Block
|
||
|
- Swish
|
||
|
File Size: 351412563
|
||
|
Tasks:
|
||
|
- Image Classification
|
||
|
ID: tf_efficientnet_b8_ap
|
||
|
LR: 0.128
|
||
|
Crop Pct: '0.954'
|
||
|
Momentum: 0.9
|
||
|
Image Size: '672'
|
||
|
Weight Decay: 1.0e-05
|
||
|
Interpolation: bicubic
|
||
|
RMSProp Decay: 0.9
|
||
|
Label Smoothing: 0.1
|
||
|
BatchNorm Momentum: 0.99
|
||
|
Code: https://github.com/rwightman/pytorch-image-models/blob/9a25fdf3ad0414b4d66da443fe60ae0aa14edc84/timm/models/efficientnet.py#L1416
|
||
|
In Collection: AdvProp
|
||
|
- Name: tf_efficientnet_b0_ap
|
||
|
Metadata:
|
||
|
FLOPs: 488688572
|
||
|
Epochs: 350
|
||
|
Batch Size: 2048
|
||
|
Training Data:
|
||
|
- ImageNet
|
||
|
Training Techniques:
|
||
|
- AdvProp
|
||
|
- AutoAugment
|
||
|
- Label Smoothing
|
||
|
- RMSProp
|
||
|
- Stochastic Depth
|
||
|
- Weight Decay
|
||
|
Architecture:
|
||
|
- 1x1 Convolution
|
||
|
- Average Pooling
|
||
|
- Batch Normalization
|
||
|
- Convolution
|
||
|
- Dense Connections
|
||
|
- Dropout
|
||
|
- Inverted Residual Block
|
||
|
- Squeeze-and-Excitation Block
|
||
|
- Swish
|
||
|
File Size: 21385973
|
||
|
Tasks:
|
||
|
- Image Classification
|
||
|
ID: tf_efficientnet_b0_ap
|
||
|
LR: 0.256
|
||
|
Crop Pct: '0.875'
|
||
|
Momentum: 0.9
|
||
|
Image Size: '224'
|
||
|
Weight Decay: 1.0e-05
|
||
|
Interpolation: bicubic
|
||
|
RMSProp Decay: 0.9
|
||
|
Label Smoothing: 0.1
|
||
|
BatchNorm Momentum: 0.99
|
||
|
Code: https://github.com/rwightman/pytorch-image-models/blob/9a25fdf3ad0414b4d66da443fe60ae0aa14edc84/timm/models/efficientnet.py#L1334
|
||
|
In Collection: AdvProp
|
||
|
Collections:
|
||
|
- Name: AdvProp
|
||
|
Paper:
|
||
|
title: Adversarial Examples Improve Image Recognition
|
||
3 years ago
|
url: https://paperswithcode.com//paper/adversarial-examples-improve-image
|
||
3 years ago
|
type: model-index
|
||
|
Type: model-index
|
||
|
-->
|