gqk/abd

Always Be Dreaming: A New Approach for Class-Incremental Learning, ICCV2021 [Unofficial PyTorch Code]

Always Be Dreaming: A New Approach for Class-Incremental Learning

An unofficial PyTorch implementation of ABD introduced in the following paper:

James Smith, Yen-Chang Hsu, Jonathan Balloch, Yilin Shen, Hongxia Jin, Zsolt Kira

Always Be Dreaming: A New Approach for Data-Free Class-Incremental Learning

International Conference on Computer Vision (ICCV), 2021.

This repo is based on cl-lite, see Experiment for usage.

The official code can be found in AlwaysBeDreaming-DFCIL.

Experiment

• Install dependencies

  pip install -r requirements.txt

• Prepare datasets

  1. create a dataset root diretory, e.g., data

  2. cifar100 will be automatically downloaded

  3. download and unzip tiny-imagenet200 to dataset root diretory

  4. follow PODNet to prepare imagenet100 dataset

  5. the overview of dataset root diretory

     ├── cifar100
     │   └── cifar-100-python
     ├── imagenet100
     │   ├── train
     │   ├── train_100.txt
     │   ├── val
     │   └── val_100.txt
     └── tiny-imagenet200
         ├── test
         ├── train
         ├── val
         ├── wnids.txt
         └── words.txt

• Generate config file (replace <root> with your dataset root path)

  python main.py --data.root <root> --print_config > cifar100.yaml

• Run experiment

  python main.py --config cifar100.yaml

We provide configs and Makefile to quickly reproduce the ten-tasks experimental results reported in the paper, run the following command if the make has been installed:

make cifar100
make tiny-imagenet200
make imagenet100

Modify fields (e.g., num_tasks) in the config files to reproduce other experiments.

Citation

@inproceedings{smith2021always,
  title={Always be dreaming: A new approach for data-free class-incremental learning},
  author={Smith, James and Hsu, Yen-Chang and Balloch, Jonathan and Shen, Yilin and Jin, Hongxia and Kira, Zsolt},
  booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
  pages={9374--9384},
  year={2021}
}

@inproceedings{gao2022rdfcil,
    title = {R-DFCIL: Relation-Guided Representation Learning for Data-Free Class Incremental Learning},
    author = {Qiankun Gao, Chen Zhao, Bernard Ghanem, Jian Zhang},
    booktitle = {European Conference on Computer Vision (ECCV)},
    year = {2022}
}