无需类别信息，通过多层级对比学习实现细粒度表征学习
Fine-Grained Representation Learning via Multi-Level Contrastive Learning without Class Priors
- 设计多层级对比损失，分离样本与特征头表示
- 在四个数据集上优于现有无监督方法，尤其在类别模糊时表现突出
- 适合缺乏标签或类别不明确的细粒度识别场景
近期无监督表征学习常依赖已知类别数以提升特征提取与聚类效果。但这一假设引发关键问题：类别数是否始终必要？类别标签能否充分捕捉数据内部细粒度特征？本文提出对比解耦（CD）框架，无需类别先验即可学习表征。CD采用多层级对比学习策略，融合实例级、特征级对比损失与归一化熵损失，以捕获语义丰富且细粒度的表示。具体而言：(1) 实例级对比损失分离不同样本的特征表示；(2) 特征级对比损失促进各特征头间的独立性；(3) 归一化熵损失保障特征多样性并防止特征坍塌。在CIFAR-10、CIFAR-100、STL-10和ImageNet-10上的大量实验表明，当类别信息缺失或模糊时，CD显著优于现有方法。代码已开源。
原文摘要 · Abstract (English)
Recent advances in unsupervised representation learning often rely on knowing the number of classes to improve feature extraction and clustering. However, this assumption raises an important question: is the number of classes always necessary, and do class labels fully capture the fine-grained features within the data? In this paper, we propose Contrastive Disentangling (CD), a framework designed to learn representations without relying on class priors. CD leverages a multi-level contrastive learning strategy, integrating instance-level and feature-level contrastive losses with a normalized entropy loss to capture semantically rich and fine-grained representations. Specifically, (1) the instance-level contrastive loss separates feature representations across samples; (2) the feature-level contrastive loss promotes independence among feature heads; and (3) the normalized entropy loss ensures feature diversity and prevents feature collapse. Extensive experiments on CIFAR-10, CIFAR-100, STL-10, and ImageNet-10 demonstrate that CD outperforms existing methods in scenarios where class information is unavailable or ambiguous. The code is available at https://github.com/Hoper-J/Contrastive-Disentangling.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。