Abstract: In recent years, various semi-supervised self-training methods have made significant progress in solving image classification problems and have attracted widespread attention. However, even the state-of-the-art methods like FixMatch can suffer from training instability and extreme performance imbalance between classes due to the accumulation of pseudo-labeling errors. To address this issue, this paper introduces the concept of robust gamma divergence on top of the FixMatch method and proposes a semi-supervised self-training approach based on gamma divergence. In this method, we incorporate gamma divergence as a regularization term to the loss function of the training model to correct the mislabeled instances caused by error accumulation. Additionally, we conduct comparative experiments on artificially polluted simulated datasets and the CIFAR-10 dataset to demonstrate the superiority of the GammaFixMatch method in handling the problem of pseudo-labeling error accumulation.

Key words: Semi-supervised, Self-training, Gamma Divergence