分离标注与预测可靠性，提升噪声标签下的模型鲁棒性
Holistic Reliability Propagation: Decoupling Annotation and Prediction for Robust Noisy-Label
- 解耦标注与伪标签的可靠性，分别建模为独立可学习参数
- 在合成与真实数据集上，平均准确率优于强基线，高噪声下仍保持竞争力
- 适合处理带噪声标签的多媒体分类任务，尤其适用于可靠性不均场景
多媒体分类中学习噪声标签时，现有方法常将外部标注与模型预测合并为单一可靠性权重，但二者失效原因不同。本文提出解耦可靠性估计：通过双层元学习为每样本生成两个批归一化标量，alpha 表示真实标签可靠性，beta 表示伪标签可靠性，且不强制两者之和为1。全貌可靠性传播（HRP）将二者分别用于不同目标：在输入分支使用基于可靠性的 Mixup 与全局门控，在对比分支采用 beta 门控的伪标签正例。在合成与真实基准上，HRP 在平均准确率上超越强基线，并在最高噪声率下保持竞争力。
原文摘要 · Abstract (English)
Learning with noisy labels in multimedia classification often combines external annotations and model predictions into a single reliability weight, even though the two sources can fail for different reasons. We instead estimate disentangled reliabilities: bilevel meta-learning produces two batch-normalized scalars per sample, alpha for the given label and beta for the pseudo-label, without constraining them to sum to one. Holistic Reliability Propagation (HRP) then routes them to different objectives, using reliability-aware Mixup with global gating on the input branch and beta-gated pseudo-label positives on the contrastive branch. On synthetic and real-world benchmarks, HRP improves average accuracy over strong baselines and remains competitive at the highest noise rates.
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。