Milking CowMask for Semi-Supervised Image Classification

TL;DR

This paper introduces CowMask, a simple mask-based augmentation technique that enhances semi-supervised image classification, achieving state-of-the-art results on ImageNet with minimal labeled data and demonstrating broad applicability across various datasets.

Contribution

The paper presents CowMask, a novel, simple augmentation method for semi-supervised learning that improves robustness and achieves top-tier results on large-scale and small-scale datasets.

Findings

State-of-the-art top-5 error of 8.76% on ImageNet with 10% labels

Competitive results on SVHN, CIFAR-10, and CIFAR-100 datasets

CowMask is simpler and effective compared to existing methods

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 state-of-the-art 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