arXiv 2403.149102024-03-22cs.CV

对抗类增量学习中的不平衡遗忘

Defying Imbalanced Forgetting in Class Incremental Learning

State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所多模态人工智能系统全国重点实验室）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；Centre for Artificial Intelligence and Robotics, HK Institute of Science & Innovation, Chinese Academy of Sciences（中国科学院香港创新研究院人工智能与机器人中心）；School of Intelligence Science and Technology, University of Science and Technology, Beijing（北京科技大学智能科学与技术学院）

Shixiong Xu, Gaofeng Meng, Xing Nie, Bolin Ni, Bin Fan, Shiming Xiang

备注：: AAAI2024

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文首次揭示了类增量学习中的不平衡遗忘现象，并提出类感知解耦（CLAD）方法，通过预测易遗忘旧类并提升其准确率，无缝集成现有方法，实现高达2.56%的性能提升。

摘要：

我们首次观察到，在同一个旧任务中，不同类别的准确率存在高度不平衡。这一在基于重放的类增量学习（CIL）中发现的有趣现象，突显了已学习类别的不平衡遗忘，因为在发生灾难性遗忘之前，它们的准确率是相似的。由于依赖平均增量准确率作为CIL的衡量标准，这一发现此前未被识别，该标准假设同一任务内的类别准确率相似。然而，在灾难性遗忘面前，这一假设是无效的。进一步的实证研究表明，这种不平衡遗忘是由语义相似的旧类和新类之间在表示上的冲突引起的。这些冲突根源于基于重放的CIL方法中存在的数据不平衡。基于这些见解，我们提出类感知解耦（CLAD），以预测更可能被遗忘的旧类并提升其准确率。重要的是，CLAD可以无缝集成到现有的CIL方法中。大量实验表明，CLAD持续改进了当前的基于重放的方法，性能提升高达2.56%。

英文摘要：

We observe a high level of imbalance in the accuracy of different classes in the same old task for the first time. This intriguing phenomenon, discovered in replay-based Class Incremental Learning (CIL), highlights the imbalanced forgetting of learned classes, as their accuracy is similar before the occurrence of catastrophic forgetting. This discovery remains previously unidentified due to the reliance on average incremental accuracy as the measurement for CIL, which assumes that the accuracy of classes within the same task is similar. However, this assumption is invalid in the face of catastrophic forgetting. Further empirical studies indicate that this imbalanced forgetting is caused by conflicts in representation between semantically similar old and new classes. These conflicts are rooted in the data imbalance present in replay-based CIL methods. Building on these insights, we propose CLass-Aware Disentanglement (CLAD) to predict the old classes that are more likely to be forgotten and enhance their accuracy. Importantly, CLAD can be seamlessly integrated into existing CIL methods. Extensive experiments demonstrate that CLAD consistently improves current replay-based methods, resulting in performance gains of up to 2.56%.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。