超越静态世界：视觉漂移下的持续类别发现
Beyond the Static World: Continual Category Discovery under Visual Drift
把部分非平衡最优传输与二值响应谱量化结合，用于无标签流式数据中已知/未知类别的自动分离；其跨域语义对齐（对抗式原型匹配）与类别拓扑一致性约束可直接迁移到开放世界识别、测试时自适应等视觉场景。
原文依据 we propose a weight-aware separation module, which leverages partial unbalanced optimal transport for instance probability modeling
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论广义类别发现（GCD）旨在借助有标签样本，从无标签数据中同时识别已知类别与新颖类别。尽管前景可观，现有大多数 GCD 方法依赖同时获取有标签与无标签数据集——这一假设在实际部署中往往并不现实。持续类别发现（CCD）通过让预训练模型适应流式无标签数据放宽了该要求，但它通常假设数据分布域一致，这严重限制了其适用性。在本工作中，我们研究开放持续类别发现（OCCD）：模型必须从可能来自异构且不断漂移的域的真实数据流中鲁棒地发现此前未见过的概念。为此，我们提出一个基于三个核心思想的自适应框架。首先，我们提出权重感知分离模块，利用部分非平衡最优传输进行实例概率建模，并采用二值响应谱量化生成用于区分已知与未知类别的线索，从而实现自动样本分离。其次，针对已知类别，我们引入跨域语义对齐模块，结合对抗学习执行自适应原型匹配，从而增强对域漂移的鲁棒性。最后，针对未知类别，我们设计类别拓扑一致性约束，在分布漂移过程中保持已知类与新颖类之间的语义关系。实验表明，我们的方法在演化域中擅长发现新类别，同时在已知类别上保持强劲性能。
英文原文摘要 1666 字符
Generalized Category Discovery (GCD) aims to identify both known and novel classes from unlabeled data with the aid of labeled examples. While promising, most existing GCD methods rely on simultaneous access to labeled and unlabeled datasets--an assumption often impractical in real-world deployments. Continual Category Discovery (CCD) relaxes this requirement by adapting a pre-trained model to streaming unlabeled data, yet it typically assumes domain-consistent data distributions. This places a strong limitation on its applicability. In this work, we study Open Continual Category Discovery (OCCD), where the model must robustly discover previously unseen concepts from real-world data streams that may originate from heterogeneous and shifting domains. To address this, we propose an adaptive framework built on three key ideas. First, we propose a weight-aware separation module, which leverages partial unbalanced optimal transport for instance probability modeling and employs binary response spectrum quantization to generate cues for distinguishing known and unknown categories, enabling automatic sample separation. Second, for known categories, we introduce a cross-domain semantic alignment module that incorporates adversarial learning to perform adaptive prototype matching, thereby enhancing robustness against domain shifts. Finally, for unknown categories, we design a category topology consistency constraint that preserves semantic relationships between known and novel classes during distribution shifts. Experiments show our approach excels at discovering new categories while maintaining strong performance on known ones in evolving domains.