mlsanityData to 14 Oct 2026

Stop Replacing Noise with Noise: Two-Source Reliability Assessment for Label Correction and Sample Reweighting in Label-Noise Learning

arXiv 2608.03432 ↗First posted 4 Aug 2026cs.LG cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Refurbishment-based noisy-label learning mixes an observed label with a model-derived pseudo target, typically using one sample-wise cleanliness score to control both branches. This creates a hidden coupling: reducing trust in the observed label automatically increases trust in the pseudo target. We show that this complementarity can replace one unreliable signal with another because a pseudo target learned from corrupted supervision may reproduce the noise it is meant to correct. Our representation diagnostics provide a consistent account of this mismatch: noisy supervision redirects deeper layers more strongly, whereas shallower relations remain comparatively stable and provide information beyond the loss posterior. We therefore propose TRACE, a Two-Source Reliability Assessment framework for Label Correction and Sample Reweighting. TRACE assesses the observed label using loss fit, shallow relation stability, and prediction agreement, while separately assessing the pseudo target using model confidence. Its source-specific scores control target correction and supervision strength without assuming complementary reliability. Across synthetic and real-world noisy benchmarks, TRACE improves representative refurbishment baselines and yields more reliable pseudo supervision.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 22 of its 31 references that are in the map. First posted on arXiv 4 Aug 2026.

Builds on

22 of 31 references.

References · 22

  1. 2018Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labelscited by 1,059
  2. 2020Prevalence of Neural Collapse during the terminal phase of deep learning trainingcited by 719
  3. 2020DivideMix: Learning with Noisy Labels as Semi-supervised Learningcited by 692
  4. 2016Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approachcited by 665
  5. 2017MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labelscited by 653
  6. 2017Robust Loss Functions under Label Noise for Deep Neural Networkscited by 443
  7. 2020Combating noisy labels by agreement: A joint training method with co-regularizationcited by 289
  8. 2017WebVision Database: Visual Learning and Understanding from Web Datacited by 262
  9. 2017Decoupling "when to update" from "how to update"cited by 230
  10. 2017CleanNet: Transfer Learning for Scalable Image Classifier Training with Label Noisecited by 181
  11. 2018Learning to Learn from Noisy Labeled Datacited by 160
  12. 2018Iterative Learning with Open-set Noisy Labelscited by 108
10 more
  1. 2021FINE Samples for Learning with Noisy Labelscited by 82
  2. 2020Error-Bounded Correction of Noisy Labelscited by 74
  3. 2022Learning with Neighbor Consistency for Noisy Labelscited by 55
  4. 2021Co-learning: Learning from Noisy Labels with Self-supervisioncited by 51
  5. 2022Mitigating Memorization of Noisy Labels by Clipping the Model Predictioncited by 40
  6. 2020Learning to Purify Noisy Labels via Meta Soft Label Correctorcited by 39
  7. 2023Learning from Noisy Labels with Decoupled Meta Label Purifiercited by 21
  8. 2022CrossSplit: Mitigating Label Noise Memorization through Data Splittingcited by 12
  9. 2023Using Unreliable Pseudo-Labels for Label-Efficient Semantic Segmentationcited by 8
  10. 2024Combating Semantic Contamination in Learning with Label Noisecited by 1

Built on by

Not cited yet.