Rebalancing Multi-Label Class-Incremental Learning

• Published in: arXiv.org
• Main Authors: Du, Kaile, Zhou, Yifan, Lyu, Fan, Li, Yuyang, Xie, Junzhou, Shen, Yixi, Hu, Fuyuan, Liu, Guangcan
• Format: Paper
• Language: English
• Published: Ithaca Cornell University Library, arXiv.org 22.08.2024
• Subjects: Labels; Learning; Performance enhancement
• ISSN: 2331-8422

Multi-label class-incremental learning (MLCIL) is essential for real-world multi-label applications, allowing models to learn new labels while retaining previously learned knowledge continuously. However, recent MLCIL approaches can only achieve suboptimal performance due to the oversight of the positive-negative imbalance problem, which manifests at both the label and loss levels because of the task-level partial label issue. The imbalance at the label level arises from the substantial absence of negative labels, while the imbalance at the loss level stems from the asymmetric contributions of the positive and negative loss parts to the optimization. To address the issue above, we propose a Rebalance framework for both the Loss and Label levels (RebLL), which integrates two key modules: asymmetric knowledge distillation (AKD) and online relabeling (OR). AKD is proposed to rebalance at the loss level by emphasizing the negative label learning in classification loss and down-weighting the contribution of overconfident predictions in distillation loss. OR is designed for label rebalance, which restores the original class distribution in memory by online relabeling the missing classes. Our comprehensive experiments on the PASCAL VOC and MS-COCO datasets demonstrate that this rebalancing strategy significantly improves performance, achieving new state-of-the-art results even with a vanilla CNN backbone.