Research map / Label-efficient and robust learning / Recognition and long-tail
Partial Label Learning with Batch Label Correction
Yan Yan, Yuhong Guo
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 |
|---|---|---|---|
| add batch label correction (ablation) | PL-BLC · Lost | accuracy: 0.616 -> 0.806 | Improved |
| add batch label correction (ablation) | PL-BLC · Yahoo! News | accuracy: 0.529 -> 0.679 | Improved |
| add MixUp data enhancement (ablation) | PL-BLC · Lost | accuracy: 0.741 -> 0.806 | Improved |
| add MixUp data enhancement (ablation) | PL-BLC · Yahoo! News | accuracy: 0.674 -> 0.679 | Improved |
| add teacher consistency loss (ablation) | PL-BLC · Lost | accuracy: 0.773 -> 0.806 | Improved |
| add teacher consistency loss (ablation) | PL-BLC · BirdSong | accuracy: 0.706 -> 0.746 | Improved |
| add all three components over classification-only baseline (ablation) | PL-BLC · Lost | accuracy: 0.562 -> 0.806 | Improved |
| add all three components over classification-only baseline (ablation) | PL-BLC · Soccer Player | accuracy: 0.470 -> 0.540 | Improved |
| add MixUp data enhancement (ablation) | PL-BLC · Soccer Player | accuracy: 0.537 -> 0.540 | No clear effect |
| use PL-BLC instead of SURE | PL-BLC · MSRCv2 | accuracy: 0.481 -> 0.536 | Improved |
| increase consistency weight eta from 0.3-0.5 to >=1 (ablation) | PL-BLC · Lost and MSRCv2 | accuracy | Hurt |
Closest papers by meaning
- Mutual Partial Label Learning with Competitive Label Noise (ICLR 2023 · similarity 0.90)
- What Makes Partial-Label Learning Algorithms Effective? (NeurIPS 2024 · similarity 0.88)
- Controller-Guided Partial Label Consistency Regularization with Unlabeled Data (AAAI 2024 · similarity 0.88)
- Distilling Reliable Knowledge for Instance-Dependent Partial Label Learning (AAAI 2024 · similarity 0.88)
- Partial Label Learning with a Partner (AAAI 2024 · similarity 0.87)
- Realistic Evaluation of Deep Partial-Label Learning Algorithms (ICLR 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.
