TALON：面向即时类别发现的测试时自适应学习

TALON: Test-time Adaptive Learning for On-the-Fly Category Discovery

March 9, 2026

作者: Yanan Wu, Yuhan Yan, Tailai Chen, Zhixiang Chi, ZiZhang Wu, Yi Jin, Yang Wang, Zhenbo Li

cs.AI

摘要

在线类别发现(OCD)旨在通过仅使用标注数据训练的模型，从未标注的在线流数据中识别已知类别并同时发现新类别。现有方法通常冻结离线训练的特征提取器，采用基于哈希的框架将特征量化为二进制码作为类别原型。然而，使用固定知识库发现新类别存在悖论，因为完全忽略了输入数据的学习潜力。此外，特征量化会导致信息损失、降低表征表达能力并加剧类内差异，常引发类别爆炸问题——即单个类别被分裂成多个伪类别。为克服这些局限，我们提出一种支持通过发现进行学习的测试时自适应框架。该框架包含两项互补策略：语义感知的原型更新和稳定的测试时编码器更新。前者动态优化类别原型以提升分类性能，后者将新信息直接整合到参数空间中。二者协同使得模型能够利用新样本持续扩展知识库。进一步地，我们在离线阶段引入边界感知逻辑校准，通过扩大类间边界并提升类内紧凑性，为未来类别发现预留嵌入空间。在标准OCD基准测试上的实验表明，本方法显著优于现有基于哈希的先进方法，在新类别识别准确率上取得明显提升，并有效缓解了类别爆炸问题。代码已公开于蓝色链接{https://github.com/ynanwu/TALON}。

English

On-the-fly category discovery (OCD) aims to recognize known categories while simultaneously discovering novel ones from an unlabeled online stream, using a model trained only on labeled data. Existing approaches freeze the feature extractor trained offline and employ a hash-based framework that quantizes features into binary codes as class prototypes. However, discovering novel categories with a fixed knowledge base is counterintuitive, as the learning potential of incoming data is entirely neglected. In addition, feature quantization introduces information loss, diminishes representational expressiveness, and amplifies intra-class variance. It often results in category explosion, where a single class is fragmented into multiple pseudo-classes. To overcome these limitations, we propose a test-time adaptation framework that enables learning through discovery. It incorporates two complementary strategies: a semantic-aware prototype update and a stable test-time encoder update. The former dynamically refines class prototypes to enhance classification, whereas the latter integrates new information directly into the parameter space. Together, these components allow the model to continuously expand its knowledge base with newly encountered samples. Furthermore, we introduce a margin-aware logit calibration in the offline stage to enlarge inter-class margins and improve intra-class compactness, thereby reserving embedding space for future class discovery. Experiments on standard OCD benchmarks demonstrate that our method substantially outperforms existing hash-based state-of-the-art approaches, yielding notable improvements in novel-class accuracy and effectively mitigating category explosion. The code is publicly available at blue{https://github.com/ynanwu/TALON}.