arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2410.06535cs.CV
收藏

Happy:一种用于持续广义类别发现的去偏学习框架

Happy: A Debiased Learning Framework for Continual Generalized Category Discovery

Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所);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, HKISI-CAS(香港中文大学(深圳)人工智能与机器人中心(中国科学院香港创新研究院));School of Computer Science and Information Engineering, Hefei University of Technology(合肥工业大学计算机与信息工程学院);School of Computer Science, University of Nottingham(诺丁汉大学计算机学院);MAIS, 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实验室)

Shijie Ma, Fei Zhu, Zhun Zhong, Wenzhuo Liu, Xu-Yao Zhang, Cheng-Lin Liu

AI总结:

针对持续广义类别发现中发现新类与防止旧类遗忘的冲突问题,提出含难度感知原型采样和软熵正则化的Happy去偏框架,有效缓解预测与难度偏差,在多数据集上取得显著性能提升。

摘要:

在不断演化的环境中,持续发现新概念至关重要。本文探讨了尚未被充分研究的持续广义类别发现(Continual Generalized Category Discovery,C-GCD)任务,该任务旨在从无标注数据中增量式发现新类别,同时保持对已学习类别的识别能力。尽管已有多种针对C-GCD任务的研究设置被提出,但它们存在局限性,无法反映真实世界场景。因此,本文研究了一种更具实用性的C-GCD设置:在更长的时间周期内需要发现更多新类别,且不存储过去类别的样本。在C-GCD任务中,模型首先在已知类别的有标注数据上进行训练,随后进入多个增量阶段,每个阶段向模型输入同时包含旧类别和新类别的无标注数据。该任务的核心挑战涉及两个相互冲突的目标:发现新类别与防止遗忘旧类别。本文深入分析了这一冲突,发现模型易受预测偏差和难度偏差的影响。为解决这些问题,本文提出了一种名为Happy的去偏学习框架,其核心特征是难度感知原型采样与软熵正则化。针对预测偏差,本文首先引入聚类引导初始化以提供鲁棒的特征表示;此外,提出软熵正则化方法为新类别分配合适的概率,可显著提升新类别的聚类性能。针对难度偏差,本文提出难度感知原型采样方法,能有效缓解模型对已见过类别的遗忘问题,尤其是困难类别的遗忘。实验结果表明,本文方法能够妥善处理C-GCD任务中的目标冲突,在多个数据集上均取得了优异性能,例如在ImageNet-100数据集上整体性能提升7.5%。本文代码已公开于https://github.com/mashijie1028/Happy-CGCD。

英文摘要:

Constantly discovering novel concepts is crucial in evolving environments. This paper explores the underexplored task of Continual Generalized Category Discovery (C-GCD), which aims to incrementally discover new classes from unlabeled data while maintaining the ability to recognize previously learned classes. Although several settings are proposed to study the C-GCD task, they have limitations that do not reflect real-world scenarios. We thus study a more practical C-GCD setting, which includes more new classes to be discovered over a longer period, without storing samples of past classes. In C-GCD, the model is initially trained on labeled data of known classes, followed by multiple incremental stages where the model is fed with unlabeled data containing both old and new classes. The core challenge involves two conflicting objectives: discover new classes and prevent forgetting old ones. We delve into the conflicts and identify that models are susceptible to prediction bias and hardness bias. To address these issues, we introduce a debiased learning framework, namely Happy, characterized by Hardness-aware prototype sampling and soft entropy regularization. For the prediction bias, we first introduce clustering-guided initialization to provide robust features. In addition, we propose soft entropy regularization to assign appropriate probabilities to new classes, which can significantly enhance the clustering performance of new classes. For the harness bias, we present the hardness-aware prototype sampling, which can effectively reduce the forgetting issue for previously seen classes, especially for difficult classes. Experimental results demonstrate our method proficiently manages the conflicts of C-GCD and achieves remarkable performance across various datasets, e.g., 7.5% overall gains on ImageNet-100. Our code is publicly available at https://github.com/mashijie1028/Happy-CGCD.

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。

↑