Home

Awesome

CoMatch: Semi-supervised Learning with Contrastive Graph Regularization, ICCV 2021 (Salesforce Research).

<img src="comatch.gif" width="600">

This is a PyTorch implementation of the <a href="https://arxiv.org/abs/2011.11183">CoMatch paper</a> <a href="https://blog.einstein.ai/comatch-advancing-semi-supervised-learning-with-contrastive-graph-regularization/">[Blog]</a>:

<pre> @inproceedings{CoMatch, title={Semi-supervised Learning with Contrastive Graph Regularization}, author={Junnan Li and Caiming Xiong and Steven C.H. Hoi}, booktitle={ICCV}, year={2021} }</pre>

Requirements:

To perform semi-supervised learning on CIFAR-10 with 4 labels per class, run:

<pre>python Train_CoMatch.py --n-labeled 40 --seed 1 </pre>

The results using different random seeds are:

seed12345avg
accuracy93.7194.1092.9390.7393.9793.09

ImageNet

For ImageNet experiments, see ./imagenet/