Research map / Label-efficient and robust learning / Noisy labels
Combating Noisy Labels with Sample Selection by Mining High-Discrepancy Examples
Xiaobo Xia, Bo Han, Yibing Zhan, Jun Yu, Mingming Gong, Chen Gong, Tongliang Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| select high-discrepancy low-loss examples (CoDis) instead of Co-teaching+ disagreement examples | Co-teaching+ · MNIST 40% symmetric noise | average selected ratio (%): 9.04 -> 60.90 | Improved |
| replace Co-teaching with CoDis on class-imbalanced asymmetric noise | Co-teaching · SVHN asymmetric 45% | test accuracy (%): 88.79 -> 94.00 | Improved |
| CoDis versus Co-teaching on balanced noise | Co-teaching · CIFAR-10 symmetric 20% | test accuracy (%): 82.35 vs 82.30 | No clear effect |
| CoDis versus JoCor on balanced NEWS with 40% symmetric noise | JoCor · NEWS symmetric 40% | test accuracy (%): 55.48 (JoCor) vs 54.93 | Hurt |
| JoCor on imbalanced noisy data | JoCor · SVHN asymmetric 45% | test accuracy (%): 67.13 vs CoDis 94.00 | Hurt |
| use CoDis sample selection inside DivideMix (DivideMix+) | DivideMix · CIFAR-10 symmetric 80% (best) | test accuracy (%): 92.29 -> 93.05 | Improved |
| use CoDis sample selection inside DivideMix | DivideMix · Clothing1M | accuracy (%): 74.76 -> 74.92 | Improved |
| CoDis versus CDR on Food-101 | CDR · Food-101 | accuracy (%): 86.36 (CDR) vs 86.13 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Co-teaching — Co-teaching: Robust training of deep neural networks with extremely noisy labels (NeurIPS 2018) · better on 6 of 7 shared results
- DivideMix — DivideMix: Learning with Noisy Labels as Semi-supervised Learning (ICLR 2020) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Coresets for Robust Training of Deep Neural Networks against Noisy Labels (NeurIPS 2020 · similarity 0.88)
- Centrality and Consistency: Two-Stage Clean Samples Identification for Learning with Instance-Dependent Noisy Labels (ECCV 2022 · similarity 0.88)
- Foster Adaptivity and Balance in Learning with Noisy Labels (ECCV 2024 · similarity 0.88)
- Regularly Truncated M-Estimators for Learning With Noisy Labels (TPAMI 2023 · similarity 0.88)
- Handling Label Noise via Instance-Level Difficulty Modeling and Dynamic Optimization (NeurIPS 2025 · similarity 0.87)
- Enhancing Sample Selection Against Label Noise by Cutting Mislabeled Easy Examples (NeurIPS 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
