Research map / Label-efficient and robust learning / Noisy labels
Instance-Dependent Label-Noise Learning With Manifold-Regularized Transition Matrix Estimation
Check 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 |
|---|---|---|---|
| Add manifold regularization to the instance-dependent transition-matrix learner | MEIDTM · Clothing1M | accuracy (%): 68.94 -> 73.05 | Helped |
| Use kernel-weighted affinity instead of the non-kernel affinity (ablation) | MEIDTM · Synthetic IDN datasets | accuracy (%): 0.06% to 2.76% | Helped |
| Use kernel-weighted affinity instead of the non-kernel affinity (ablation) | MEIDTM · Clothing1M | accuracy (%): 73.05 -> 73.34 | Helped |
| Integrate kMEIDTM with DivideMix | DivideMix · Clothing1M | accuracy (%): +0.15% | Helped |
| Integrate kMEIDTM with DivideMix | DivideMix · Food101N | accuracy (%): +1.22% | Helped |
| Increase the manifold-regularization weight λ (ablation) | MEIDTM · CIFAR-10 IDN-10% versus IDN-50% | accuracy (%) | Mixed |
| Set λ around 0.3 (ablation) | MEIDTM · CIFAR-10, five IDN noise rates | accuracy (%) | Helped |
| Train the model for more epochs while optimizing the instance-dependent transition matrix (ablation) | MEIDTM · CIFAR-10 IDN, five noise rates | transition-matrix estimation error (L1) | Helped |
Closest papers by meaning
- Part-dependent Label Noise: Towards Instance-dependent Label Noise (NeurIPS 2020 · similarity 0.92)
- Identifiability of Label Noise Transition Matrix (ICML 2023 · similarity 0.92)
- Learning Causal Transition Matrix for Instance-dependent Label Noise (AAAI 2025 · similarity 0.90)
- Learning the Latent Causal Structure for Modeling Label Noise (NeurIPS 2024 · similarity 0.89)
- Estimation of the Label-Noise Transition Matrix with Performance Guarantees via Selective Classification (NeurIPS 2026 · similarity 0.88)
- Estimating Noisy Class Posterior with Part-level Labels for Noisy Label Learning (CVPR 2024 · 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.