Research map / Label-efficient and robust learning / Noisy labels
Estimating Noise Transition Matrix with Label Correlations for Noisy Multi-Label Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 unbiased (gold) sample selection instead of GMM small-loss selection (ablation) | Our estimator · Pascal-VOC2007 noise (0,0.6) | estimation error: 2.30 -> 0.51 | Helped |
| use label-correlation estimator instead of Dual T-estimator max | transition matrix estimation · Pascal-VOC2012 noise (0,0.6) | estimation error: 5.13 -> 1.94 | Helped |
| use label-correlation estimator instead of Dual T-estimator max | transition matrix estimation · MS-COCO noise (0,0.2) | estimation error: 5.04 -> 7.42 | Hurt |
| reweighting with estimated transition matrix instead of standard training | ResNet-50 multi-label classifier · MS-COCO noise (0.6,0) | CF1: 7.07 -> 58.63 | Helped |
| use our estimator for Reweight instead of Reweight-DualT 97% | Reweight · Pascal-VOC2007 noise (0.6,0) | mAP: 75.76 -> 74.46 | Hurt |
Closest papers by meaning
- Dual T: Reducing Estimation Error for Transition Matrix in Label-noise Learning (NeurIPS 2020 · similarity 0.94)
- Estimation of the Label-Noise Transition Matrix with Performance Guarantees via Selective Classification (NeurIPS 2026 · similarity 0.92)
- Learning the Latent Causal Structure for Modeling Label Noise (NeurIPS 2024 · similarity 0.91)
- Estimating Noisy Class Posterior with Part-level Labels for Noisy Label Learning (CVPR 2024 · similarity 0.90)
- Learning Causal Transition Matrix for Instance-dependent Label Noise (AAAI 2025 · similarity 0.89)
- Dirichlet-based Per-Sample Weighting by Transition Matrix for Noisy Label Learning (ICLR 2024 · similarity 0.88)
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.