Research map / Label-efficient and robust learning / Noisy labels
MTaDCS: Moving Trace and Feature Density-based Confidence Sample Selection under Label Noise
Qingzheng Huang, Xilin He, Xiaole Xian, Qinliang Lin, Weicheng Xie, Siyang Song, Linlin Shen, Zitong Yu
This paper tried a common change: Add a contrastive loss
add local contrastive learning on the confidence set · CIFAR-10 noisy labels · feature clustering / sample selection Improved
What 1,317 accepted papers report for this change
- 1,253 improved
- 22 no clear effect
- 16 mixed
- 26 hurt
- Beyond Looking Up, Try Looking Around: Harmonizing Global Structure and Local Consistency in Optimal Transport for Short Text Clustering ICML 2026 · add contrastive learning · SearchSnippets · ACC: 56.65 -> 80.12Improved
- SleepLM: Natural-Language Intelligence for Human Sleep ICML 2026 · add contrastive objective to captioning (CoCa vs Cap) · sleep event IoU · IoU: 29.2 -> 27.4Hurt
- LIMSSR: LLM-Driven Sequence-to-Score Reasoning under Training-Time Incomplete Multimodal Observations ICML 2026 · add contrastive loss L_con · FS1000 average · Spearman correlation: 0.759 -> 0.789Improved
See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.
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 |
|---|---|---|---|
| use LOF-based local sample selection in class-wise feature space (ablation) | MTaDCS · CIFAR-10 symmetric noise 50% | test accuracy (%): 59.1 -> 83.2 | Improved |
| add confidence set and non-confidence set on top of LOF (ablation) | MTaDCS · CIFAR-10 symmetric noise 50% | test accuracy (%): 83.2 -> 90.9 | Improved |
| add confidence set on top of LOF (ablation) | MTaDCS · CIFAR-10 symmetric noise 90% | test accuracy (%): 76.7 -> 62.0 | Hurt |
| add mislabeled-sample removal from the confidence set (ablation) | MTaDCS · CIFAR-10 symmetric noise 90% | test accuracy (%): 62.0 -> 77.3 | Improved |
| add Mixup on the confidence set (ablation) | MTaDCS · CIFAR-10 symmetric noise 50% | test accuracy (%) | Improved |
| add self-ensemble label correction (ablation) | MTaDCS · CIFAR/noisy-label settings | test accuracy (%) | Improved |
| add local contrastive learning on the confidence set (ablation) | MTaDCS · CIFAR-10 noisy labels | feature clustering / sample selection | Improved |
| use MTaDCS instead of TCL on CIFAR-100 90% symmetric noise | MTaDCS · CIFAR-100 symmetric noise 90% | test accuracy (%): 54.5 -> 61.4 | Improved |
| use MTaDCS instead of RRL on CIFAR-100 20% symmetric noise | MTaDCS · CIFAR-100 symmetric noise 20% | test accuracy (%): 79.1 -> 77.1 | Hurt |
| use MTaDCS instead of TCL on WebVision-50 top-1 | MTaDCS · WebVision-50 val | top-1 accuracy (%): 79.1 -> 79.1 | No clear effect |
Closest papers by meaning
- Enhanced Sample Selection with Confidence Tracking: Identifying Correctly Labeled Yet Hard-to-Learn Samples in Noisy Data (AAAI 2025 · similarity 0.90)
- Estimation of the Label-Noise Transition Matrix with Performance Guarantees via Selective Classification (NeurIPS 2026 · similarity 0.85)
- NoiseGPT: Label Noise Detection and Rectification through Probability Curvature (NeurIPS 2024 · similarity 0.85)
- Enhancing Sample Selection Against Label Noise by Cutting Mislabeled Easy Examples (NeurIPS 2025 · similarity 0.85)
- Pearls from Pebbles: Improved Confidence Functions for Auto-labeling (NeurIPS 2024 · similarity 0.85)
- Regroup Median Loss for Combating Label Noise (AAAI 2024 · similarity 0.85)
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.
