标题： L3A：多标签类增量学习中的标签增强分析适应 显示英文标题

标题： L3A: Label-Augmented Analytic Adaptation for Multi-Label Class Incremental Learning

摘要： 类增量学习（CIL）使模型能够持续学习新类别，同时不会忘记之前学到的知识。多标签类增量学习（MLCIL）将CIL扩展到现实场景中，其中每个样本可能属于多个类别，这带来了几个挑战：由于缺失标签导致的标签缺失问题，以及由于类别不平衡引起的模型偏向多数类别问题。 为了解决这些挑战，我们提出了Label-Augmented Analytic Adaptation（L3A），这是一种无需存储过去样本的无范例方法。L3A集成了两个关键模块。伪标签（PL）模块通过为当前阶段样本生成伪标签来实现标签增强，从而解决标签缺失问题。加权分析分类器（WAC）为神经网络推导出一个闭合形式的解决方案。它引入了样本特定的权重，以自适应地平衡类别贡献并缓解类别不平衡问题。 MS-COCO和PASCAL VOC数据集上的实验表明，L3A在MLCIL任务中优于现有方法。我们的代码可在https://github.com/scut-zx/L3A获取。 显示英文摘要

摘要： Class-incremental learning (CIL) enables models to learn new classes continually without forgetting previously acquired knowledge. Multi-label CIL (MLCIL) extends CIL to a real-world scenario where each sample may belong to multiple classes, introducing several challenges: label absence, which leads to incomplete historical information due to missing labels, and class imbalance, which results in the model bias toward majority classes. To address these challenges, we propose Label-Augmented Analytic Adaptation (L3A), an exemplar-free approach without storing past samples. L3A integrates two key modules. The pseudo-label (PL) module implements label augmentation by generating pseudo-labels for current phase samples, addressing the label absence problem. The weighted analytic classifier (WAC) derives a closed-form solution for neural networks. It introduces sample-specific weights to adaptively balance the class contribution and mitigate class imbalance. Experiments on MS-COCO and PASCAL VOC datasets demonstrate that L3A outperforms existing methods in MLCIL tasks. Our code is available at https://github.com/scut-zx/L3A.