Research map / Label-efficient and robust learning / Noisy labels
Knockoffs-SPR: Clean Sample Selection in Learning With Noisy Labels
Yikai Wang, Yanwei Fu, Xinwei Sun
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 |
|---|---|---|---|
| use most-confident label permutation instead of random permutation for the knockoff copy (ablation) | Knockoffs-SPR · CIFAR-10 Sym. 40% | accuracy (%): 92.0 -> 94.7 | Improved |
| use most-confident label permutation instead of random permutation (ablation) | Knockoffs-SPR · CIFAR-10 Asy. 40% | accuracy (%): 84.4 -> 93.5 | Improved |
| use most-confident label permutation instead of random permutation (ablation) | Knockoffs-SPR · CIFAR-10 Sym. 80% | FSR (%): 49.76 -> 26.72 | Improved |
| use most-confident label permutation instead of random permutation (ablation) | Knockoffs-SPR · CIFAR-10 Sym. 80% | accuracy (%): 84.6 -> 84.3 | No clear effect |
| add class-specific selection (ablation) | Knockoffs-SPR · CIFAR-10 Sym. 80% | accuracy (%): 83.0 -> 84.3 | Improved |
| add class-specific selection (ablation) | Knockoffs-SPR · CIFAR-10 Sym. 40% | accuracy (%): 94.4 -> 94.7 | Improved |
| add PCA preprocessing of features (ablation) | Knockoffs-SPR · CIFAR-10 Sym. 80% | accuracy (%): 10.0 -> 84.3 | Improved |
| add PCA preprocessing of features (ablation) | Knockoffs-SPR · CIFAR-10 Sym. 40% | accuracy (%): 81.7 -> 94.7 | Improved |
| add PCA preprocessing of features (ablation) | Knockoffs-SPR · CIFAR-10 Asy. 40% | accuracy (%): 93.7 -> 93.5 | No clear effect |
| use PCA instead of an extra linear layer for dimension reduction (ablation) | Knockoffs-SPR · CIFAR-10 Sym. 80% | accuracy (%): 79.3 -> 84.3 | Improved |
| use Knockoffs-SPR instead of SPR | SPR · CIFAR-10 Sym. 80% | accuracy (%): 78.0 -> 84.3 | Improved |
| use Knockoffs-SPR instead of SPR | SPR · CIFAR-100 Sym. 80% | accuracy (%): 38.1 -> 30.5 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DivideMix — DivideMix: Learning with Noisy Labels as Semi-supervised Learning (ICLR 2020) · better on 1 of 1 shared result
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
- Scalable Penalized Regression for Noise Detection in Learning With Noisy Labels (CVPR 2022 · similarity 0.91)
- Self-Filtering: A Noise-Aware Sample Selection for Label Noise with Confidence Penalization (ECCV 2022 · similarity 0.89)
- Enhancing Sample Selection Against Label Noise by Cutting Mislabeled Easy Examples (NeurIPS 2025 · similarity 0.88)
- Enhanced Sample Selection with Confidence Tracking: Identifying Correctly Labeled Yet Hard-to-Learn Samples in Noisy Data (AAAI 2025 · similarity 0.87)
- Sample Selection via Contrastive Fragmentation for Noisy Label Regression (NeurIPS 2024 · similarity 0.87)
- Towards a statistical theory of data selection under weak supervision (ICLR 2024 · similarity 0.86)
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.
