Skip to content

robot-vision-454P/FSR_GAN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Model

생성모델을 사용하여 non-robust feature에서 원래 feature로 생성할 수 있도록 하였다.

생성 모델을 활용하면, non-robust feature을 original feature와 비슷하게 생성하여 적대적 예제에도 강건한 모델을 생성할 수 있다.

MODEL_FIGURE

Setup

conda env create -n FSRGAN python=3.7
conda activate FSRGAN
conda install pytorch==1.12.1 torchvision==0.13.1 torchaudio==0.12.1 cudatoolkit=11.3 -c pytorch
conda install tqdm

Train

python train.py --save_name cifar10_vgg16 --dataset cifar10 --model vgg16 --device 0 --epoch 200 --bs 256

Test

python test.py --load_name cifar10_vgg16 --dataset cifar10 --model vgg16 --device 0

Reference

https://github.com/wkim97/FSR

https://github.com/sky4689524/DefenseGAN-Pytorch

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages