改进图像分类损失函数，让模型更懂类别之间的上下级关系。
When Labels Have Structure: Improving Image Classification with Hierarchy-Aware Cross-Entropy
- 用层级结构重构交叉熵损失，让父类继承子类的置信度。
- 在18组实验中平均提升4.66%准确率，线性探测下平均增益2.18%。
- 适合有明确类别层次的数据集，如细粒度图像分类任务。
标准交叉熵损失在几乎所有机器学习任务中被广泛使用，但它对所有误分类一视同仁，忽略了类别层级所蕴含的语义距离。本文提出层级感知交叉熵（HACE），一种可直接替代标准交叉熵的损失函数，将已知的类别层级结构融入训练过程。HACE包含两个组件：预测聚合，将模型概率向上传播至父类，使父节点累积子类的置信度；祖先标签平滑，将真实标签信号沿从真类到根节点的路径分布。我们在CIFAR-100、FGVC Aircraft和NABirds三个数据集上评估HACE，涵盖六种不同架构（含卷积与注意力结构）的端到端训练，以及基于冻结DINOv2-Large特征的线性探测。在端到端训练中，HACE在18组实验中的15组超越标准交叉熵，平均提升4.66%；在线性探测中，其在三个数据集上均优于所有对比方法，平均比次优基线高出2.18%。
原文摘要 · Abstract (English)
Standard cross-entropy is the default classification loss across virtually all of machine learning, yet it treats all misclassifications equally, ignoring the semantic distances that a class hierarchy encodes. We propose Hierarchy-Aware Cross-Entropy (HACE), a drop-in replacement for standard cross-entropy that incorporates a known class hierarchy directly into the loss. HACE combines two components: prediction aggregation, which propagates the model's probability mass upward through the class hierarchy to ensure that parent nodes accumulate the confidence of their children; and ancestral label smoothing, which distributes the ground-truth signal along the path from the true class to the root. We evaluate HACE on CIFAR-100, FGVC Aircraft, and NABirds in two regimes: end-to-end training across six architectures spanning convolutional and attention-based designs, and linear probing on frozen DINOv2-Large features. In end-to-end training, HACE improves accuracy over standard cross-entropy in 15 out of 18 architecture--dataset pairs, with a mean gain of 4.66\%. In linear probing on frozen DINOv2-Large features, HACE outperforms all competing methods on all three datasets, with a mean improvement of 2.18\% over the next best baseline.
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。