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

arXiv: https://arxiv.org/abs/2608.03432 · HTML: https://mlsanity.com/paper/2608.03432/

## 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 (most cited first, 22 of 22)

- [Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels](https://mlsanity.com/paper/1804.06872/) (2018)
- [Prevalence of Neural Collapse during the terminal phase of deep learning training](https://mlsanity.com/paper/2008.08186/) (2020)
- [DivideMix: Learning with Noisy Labels as Semi-supervised Learning](https://mlsanity.com/paper/2002.07394/) (2020)
- [Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approach](https://mlsanity.com/paper/1609.03683/) (2016)
- [MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels](https://mlsanity.com/paper/1712.05055/) (2017)
- [Robust Loss Functions under Label Noise for Deep Neural Networks](https://mlsanity.com/paper/1712.09482/) (2017)
- [Combating noisy labels by agreement: A joint training method with co-regularization](https://mlsanity.com/paper/2003.02752/) (2020)
- [WebVision Database: Visual Learning and Understanding from Web Data](https://mlsanity.com/paper/1708.02862/) (2017)
- [Decoupling "when to update" from "how to update"](https://mlsanity.com/paper/1706.02613/) (2017)
- [CleanNet: Transfer Learning for Scalable Image Classifier Training with Label Noise](https://mlsanity.com/paper/1711.07131/) (2017)
- [Learning to Learn from Noisy Labeled Data](https://mlsanity.com/paper/1812.05214/) (2018)
- [Iterative Learning with Open-set Noisy Labels](https://mlsanity.com/paper/1804.00092/) (2018)
- [FINE Samples for Learning with Noisy Labels](https://mlsanity.com/paper/2102.11628/) (2021)
- [Error-Bounded Correction of Noisy Labels](https://mlsanity.com/paper/2011.10077/) (2020)
- [Learning with Neighbor Consistency for Noisy Labels](https://mlsanity.com/paper/2202.02200/) (2022)
- [Co-learning: Learning from Noisy Labels with Self-supervision](https://mlsanity.com/paper/2108.04063/) (2021)
- [Mitigating Memorization of Noisy Labels by Clipping the Model Prediction](https://mlsanity.com/paper/2212.04055/) (2022)
- [Learning to Purify Noisy Labels via Meta Soft Label Corrector](https://mlsanity.com/paper/2008.00627/) (2020)
- [Learning from Noisy Labels with Decoupled Meta Label Purifier](https://mlsanity.com/paper/2302.06810/) (2023)
- [CrossSplit: Mitigating Label Noise Memorization through Data Splitting](https://mlsanity.com/paper/2212.01674/) (2022)
- [Using Unreliable Pseudo-Labels for Label-Efficient Semantic Segmentation](https://mlsanity.com/paper/2306.02314/) (2023)
- [Combating Semantic Contamination in Learning with Label Noise](https://mlsanity.com/paper/2412.11620/) (2024)

## 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.