提出新方法诊断标签噪声，更好区分真实样本与错误标签。
Radial-Angular Geometry for Reliable Update Diagnosis in Noisy-Label Learning
- 用梯度冲突检测机制，比较样本与教师模型的更新方向
- 在多个数据集上提升难样本保留率和模型准确率
- 适合标签质量差的场景，如真实世界数据标注
现有噪声标签方法通常通过损失、置信度或熵等前向信号估计样本可靠性，但这些信号仅反映预测难度，无法判断标签是否引发可靠参数更新。本文将可靠性评估重构为对标签更新的诊断问题。样本级经验费雪迹提供反向空间的更新能量度量，在分类层可分解为预测残差项与特征敏感性项，捕捉比标量损失更丰富的信息。然而，迹仍为径向幅值信号，无法判断大更新是有益还是有害。为此，本文提出相对几何冲突（RGC），通过对比观察标签梯度与基于指数移动平均教师模型生成的参考梯度，识别更新方向的一致性。该冲突项能有效区分大而一致的困难干净样本更新与因标签污染导致的大冲突更新。在合成及真实世界噪声标签基准测试中，RGC显著提升了困难干净样本的保留率与整体准确率。
原文摘要 · Abstract (English)
Noisy-label methods often estimate sample reliability from forward-space signals such as loss, confidence, or entropy. These signals indicate whether a sample is difficult to predict, but they do not directly test whether its observed label induces a reliable parameter update. This gap matters because hard clean samples and mislabeled samples can have similar loss while inducing different updates. We recast reliability estimation as diagnosis of the observed-label update. The sample-wise empirical Fisher trace gives a backward-space measure of update energy: for the classifier layer, it factorizes into a prediction-residual term and a feature-sensitivity term, so it captures information beyond scalar loss. Trace, however, is still a radial magnitude signal and cannot decide whether a large update is useful or harmful. We therefore propose Relative Geometric Conflict (RGC), which compares the observed-label gradient with a reference gradient induced by an EMA teacher. The conflict term helps distinguish large but aligned hard-clean updates from large conflicting updates caused by corrupted labels. Across synthetic and real-world noisy-label benchmarks, RGC improves hard-clean preservation and accuracy under our evaluation protocol.
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。