FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence

Semi-supervised learning (SSL) provides an effectivemeans of leveraging unlabeled data to improve a model’sperformance. In this paper, we demonstrate the power of asimple combination of two common SSL methods: consistency regularization and pseudo-labeling. Our algorithm,FixMatch, first generates pseudo-labels using the model’spredictions on weakly-augmented unlabeled images. For agiven image, the pseudo-label is only retained if the modelproduces a high-confidence prediction. The model is thentrained to predict the pseudo-label when fed a stronglyaugmented version of the same image. Despite its simplicity, we show that FixMatch achieves state-of-the-art performance across a variety of standard semi-supervised learning benchmarks, including 94.93% accuracy on CIFAR-10with 250 labels and 88.61% accuracy with 40 – just 4 labels per class. Since FixMatch bears many similaritiesto existing SSL methods that achieve worse performance,we carry out an extensive ablation study to tease apartthe experimental factors that are most important to FixMatch’s success. We make our code available at https://github.com/google-research/fixmatch.