Skip to content

Latest commit

 

History

History
31 lines (26 loc) · 715 Bytes

README.md

File metadata and controls

31 lines (26 loc) · 715 Bytes

Memory-Free Generative Replay For Class-Incremental Learning

Description

This is the source code for "Memory-Free Generative Replay For Class-Incremental Learning"

Getting Started

Dependencies

  • Python 3.7
  • Pytorch 1.6.0
  • Sklearn

the environment can be created by utilizing MFGR.yml

conda env create -f MFGR.yml

Run the Code

We employ CIFAR-100 as an example

sh run.sh

Citation

@article{xin2021memory,
  title={Memory-Free Generative Replay For Class-Incremental Learning},
  author={Xin, Xiaomeng and Zhong, Yiran and Hou, Yunzhong and Wang, Jinjun and Zheng, Liang},
  journal={arXiv preprint arXiv:2109.00328},
  year={2021}
}