SODA

The implementation of SODA on CIFAR-10-C, CIFAR-100-C and ImageNet-C.

Prerequisites:

• python == 3.10.8
• cudatoolkit == 11.7
• pytorch ==1.13.1
• torchvision == 0.14.1
• numpy, PIL, argparse, collections, math, random

Datasets

Please download and organize CIFAR-10-C, CIFAR-100-C and ImageNet-C in this structure:

(ImageNet-C data can also be generated following instructions in this repository)

BETA
├── data
    ├──CIFAR-10
    │   ├── CIFAR-10-C
    │   │   ├── brightness.npy
    │   │   ├── contrast.npy
    │   │   ├── ...
    │   │   ├── labels.npy
    ├──CIFAR-100
    │   ├── CIFAR-100-C
    │   │   ├── brightness.npy
    │   │   ├── contrast.npy
    │   │   ├── ...
    │   │   ├── labels.npy
    ├──ImageNet
    │   ├── ImageNet-C
    │   │   ├── brightness.pth
    │   │   ├── contrast.pth
    │   │   ├── ...
    │   │   ├── labels.pth

Pre-trained Models

The checkpoints of pre-trained Resnet-50 can be downloaded (197MB) using the following command:

mkdir -p results/cifar10_joint_resnet50 && cd results/cifar10_joint_resnet50
gdown https://drive.google.com/uc?id=1MZN19o-5b2w-BI1ObIlnsJ8XBZvMuL77 && cd ../..
mkdir -p results/cifar100_joint_resnet50 && cd results/cifar100_joint_resnet50
gdown https://drive.google.com/uc?id=1C7knE2S9kKDYZrqd4Bo4S5lOgp7Le_DP && cd ../..
mkdir -p results/imagenet && cd results/imagenet
gdown https://drive.google.com/uc?id=1GSGzOv0MNBBMEYeRRQlp1WGD1USDl0iP && cd ../..

The CIFAR-10/100 pre-trained models are obtained by training on the clean CIFAR-10/100 images using semi-supervised SimCLR. The ImageNet pre-trained model is obtained from TorchVision

Adaptation on CIFAR-10-C

# offline SODA
bash scripts/run_offline_soda_10.sh

# offline SODA-R
bash scripts/run_offline_soda_r_10.sh

# offline MA-SO
bash scripts/run_offline_ma_10.sh

# online SODA-O
bash scripts/run_online_soda_10.sh

Adaptation on CIFAR-100-C

# offline SODA
bash scripts/run_offline_soda_100.sh

# offline SODA-R
bash scripts/run_offline_soda_r_100.sh

# offline MA-SO
bash scripts/run_offline_ma_100.sh

# online SODA-O
bash scripts/run_online_soda_100.sh

Adaptation on ImageNet-C

# offline SODA
bash scripts/run_offline_soda_imagenet.sh

# offline SODA-R
bash scripts/run_offline_soda_r_imagenet.sh

# offline MA-SO
bash scripts/run_offline_ma_imagenet.sh