Holistic Reliability Propagation: Decoupling Annotation and Prediction for Robust Noisy-Label
Year: 2026
Type: preprint
Abstract: 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.
Source: arXiv (Cornell University)
Authors: Mao, Jingyang, Peng, Ningkang, Gu, Yanhui
Institution: Nanjing Normal University
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7162014649
Cites:
Cited by:
Related to: