This is the implementation code of CVPR 2022 paper 'Federated Class-Incremental Learning'.
python == 3.6
torch == 1.2.0
numpy
PIL
torchvision == 0.4.0
cv2
scipy == 1.5.2
sklearn == 0.24.1
You don't need to do anything before running the experiments of CIFAR100.
You need to download the Mini-Imagenet from here and place it in './train'.
You need to download the Tiny-Imagenet from here and place it in './tiny-imagenet-200'.
python fl_main.py
The detailed arguments can be found in './src/option.py'.
If you find our work is helpful to your research, please consider to cite.
@InProceedings{dong2022federated,
author = {Jiahua Dong and Lixu Wang and Zhen Fang and Gan Sun and Shichao Xu and Xiao Wang and Qi Zhu},
title = {Federated Class Incremental Learning},
booktitle = {CVPR},
year = {2022}
}