Hier-COS：通过正交子空间组合使深度特征具备层次感知能力
Hier-COS: Making Deep Features Hierarchy-aware via Composition of Orthogonal Subspaces
Hier-COS 提供了一种将预训练骨干网络（如 ViT）的冻结特征转换为层次感知表征的通用框架，其正交子空间组合机制可被视觉研究者复用于需要结构化语义先验的细粒度分类任务。
原文依据 Hier-COS can effectively learn to transform the frozen features extracted from a pretrained backbone (ViT) to be hierarchy-aware
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论传统分类器将所有类别标签视为相互独立，从而认为所有负类都是同等错误的。这种方法在许多现实场景中严重失效，因为已知的语义层次结构定义了负类之间的偏好偏序。虽然层次感知的特征表示已被证明有望缓解这一问题，但其性能通常使用错误严重度（MS）和平均层次距离（AHD）等指标进行评估。本文强调了现有层次评估指标的重要缺陷，证明它们往往无法衡量真实的层次性能。我们的分析表明，尽管现有方法在MS和AHD分数上具有竞争力，但它们学习到的是次优的层次表示。为解决这些问题，我们引入了正交子空间层次组合（Hier-COS），这是一个用于统一“层次感知细粒度”和“层次多标签”分类的新框架。我们证明Hier-COS在理论上保证与给定的层次树一致。此外，我们的框架根据类别在层次树中的位置隐式地调整不同类别的学习能力——这是现有方法所缺乏的重要特性。最后，为解决评估指标的局限性，我们提出了层次有序偏好分数（HOPS），这是一种基于排序的指标，可证明克服了当前评估标准的缺陷。我们在四个具有挑战性的数据集上对Hier-COS进行了基准测试，包括深度且不平衡的tieredImageNet-H（12层）和iNaturalist-19（7层）。通过大量实验，我们证明Hier-COS在每个数据集的所有层次指标上均达到最优性能，同时在除一个案例外的所有情况下均优于top-1准确率。最后，我们表明Hier-COS能够有效地学习将预训练骨干网络（ViT）提取的冻结特征转换为层次感知特征，从而为层次分类性能带来显著收益。
英文原文摘要 2119 字符
Traditional classifiers treat all class labels as mutually independent, thereby considering all negative classes to be equally incorrect. This approach fails severely in many real-world scenarios, where a known semantic hierarchy defines a partial order of preferences over negative classes. While hierarchy-aware feature representations have shown promise in mitigating this problem, their performance is typically assessed using metrics like Mistake Severity (MS) and Average Hierarchical Distance (AHD). In this paper, we highlight important shortcomings in existing hierarchical evaluation metrics, demonstrating that they are often incapable of measuring true hierarchical performance. Our analysis reveals that existing methods learn sub-optimal hierarchical representations, despite competitive MS and AHD scores. To counter these issues, we introduce Hierarchical Composition of Orthogonal Subspaces (Hier-COS), a novel framework for unified 'hierarchy-aware fine-grained' and 'hierarchical multi-label' classification. We show that Hier-COS is theoretically guaranteed to be consistent with the given hierarchy tree. Furthermore, our framework implicitly adapts the learning capacity for different classes based on their position within the hierarchy tree -- a vital property absent in existing methods. Finally, to address the limitations of evaluation metrics, we propose Hierarchically Ordered Preference Score (HOPS), a ranking-based metric that demonstrably overcomes the deficiencies of current evaluation standards. We benchmark Hier-COS on four challenging datasets, including the deep and imbalanced tieredImageNet-H (12-level) and iNaturalist-19 (7-level). Through extensive experiments, we demonstrate that Hier-COS achieves state-of-the-art performance across all hierarchical metrics for every dataset, while simultaneously beating the top-1 accuracy in all but one case. Lastly, we show that Hier-COS can effectively learn to transform the frozen features extracted from a pretrained backbone (ViT) to be hierarchy-aware, yielding substantial benefits for hierarchical classification performance.