CoSSL: Co-Learning of Representation and Classifier for Imbalanced Semi-Supervised Learning

Standard semi-supervised learning (SSL) using classbalanced datasets has shown great progress to leverage unlabeled data effectively. However, the more realistic setting of class-imbalanced data -called imbalanced SSLis largely underexplored and standard SSL tends to underperform. In this paper, we propose a novel co-learning framework (CoSSL), which decouples representation and classifier learning while coupling them closely. To handle the data imbalance, we devise Tail-class Feature Enhancement (TFE) for classifier learning. Furthermore, the current evaluation protocol for imbalanced SSL focuses only on balanced test sets, which has limited practicality in realworld scenarios. Therefore, we further conduct a comprehensive evaluation under various shifted test distributions. In experiments, we show that our approach outperforms other methods over a large range of shifted distributions, achieving state-of-the-art performance on benchmark datasets ranging from CIFAR-10, CIFAR-100, ImageNet, to Food-101. Code is available at https://github. com/YUE-FAN/CoSSL.