Milking CowMask for Semi-supervised Image Classification

Geoff French, Geoff French, Avital Oliver, Tim Salimans

2022

Abstract

Consistency regularization is a technique for semi-supervised learning that underlies a number of strong results for classification with few labeled data. It works by encouraging a learned model to be robust to perturbations on unlabeled data. Here, we present a novel mask-based augmentation method called CowMask. Using it to provide perturbations for semi-supervised consistency regularization, we achieve a competitive result on ImageNet with 10% labeled data, with a top-5 error of 8.76% and top-1 error of 26.06%. Moreover, we do so with a method that is much simpler than many alternatives. We further investigate the behavior of CowMask for semi-supervised learning by running many smaller scale experiments on the SVHN, CIFAR-10 and CIFAR-100 data sets, where we achieve results competitive with the state of the art, indicating that CowMask is widely applicable. We open source our code at https://github.com/google-research/google-research/tree/master/milking cowmask.

Paper Citation

in Harvard Style

French G., Oliver A. and Salimans T. (2022). Milking CowMask for Semi-supervised Image Classification. In Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP; ISBN 978-989-758-555-5, SciTePress, pages 75-84. DOI: 10.5220/0010773700003124

in Bibtex Style

@conference{visapp22,
author={Geoff French and Avital Oliver and Tim Salimans},
title={Milking CowMask for Semi-supervised Image Classification},
booktitle={Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP},
year={2022},
pages={75-84},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010773700003124},
isbn={978-989-758-555-5},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP
TI - Milking CowMask for Semi-supervised Image Classification
SN - 978-989-758-555-5
AU - French G.
AU - Oliver A.
AU - Salimans T.
PY - 2022
SP - 75
EP - 84
DO - 10.5220/0010773700003124
PB - SciTePress