Boosting Semi-supervised Learning by Exploiting All Unlabeled Data
Yuhao Chen, Xin Tan, Borui Zhao, Zhaowei Chen, Renjie Song, Jiajun Liang, Xuequan Lu
CVPR 2023, Arxiv
This repo is the Megengine implementation of FullMatch.
-
Install MegEngine (version==1.12.2/1.12.3)
-
For training FullMatch:
python fullmatch.py --c config/fullmatch/fullmatch_cifar100.yaml
- For training FullFlex:
python fullflex.py --c config/fullflex/fullflex_cifar100.yaml
Since the official Megengine does not support many classification benchmarks (e.g., SVHN, STL10)
We thanks the TorchSSL project for reference.
All origin train logs and models are in this link (pytorch framework), and this code can produce the similar performance based on megengine framework.
FullMatch is released under the Apache 2.0 license. See LICENSE for details.
@inproceedings{chen2023boosting,
title={Boosting Semi-Supervised Learning by Exploiting All Unlabeled Data},
author={Chen, Yuhao and Tan, Xin and Zhao, Borui and Chen, Zhaowei and Song, Renjie and Liang, Jiajun and Lu, Xuequan},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={7548--7557},
year={2023}
}