Research map / Label-efficient and robust learning / Noisy labels
Combating Noisy Labels by Agreement: A Joint Training Method with Co-Regularization
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 |
|---|---|---|---|
| Use JoCoR instead of Co-teaching+ | two-network noisy-label training · MNIST Symmetry-80% | test accuracy: 58.92 -> 84.89 | Helped |
| Use JoCoR instead of Co-teaching+ | two-network noisy-label training · CIFAR-10 Symmetry-20% | test accuracy: 78.71 -> 85.73 | Helped |
| Use JoCoR instead of Co-teaching+ | two-network noisy-label training · CIFAR-100 Asymmetry-40% | test accuracy: 33.62 -> 32.70 | Hurt |
| Use JoCoR instead of Co-teaching | noisy-label learning on Clothing1M · Clothing1M test set | classification accuracy (%) last: 68.51 -> 69.79 | Helped |
| Add Co-Regularization to Joint-only training (ablation) | Joint-only · MNIST Symmetry-50% | test accuracy | Helped |
| Add Co-Regularization to Joint-only training (ablation) | Joint-only · CIFAR-10 Symmetry-50% | test accuracy | Helped |
| Use Joint Training rather than cross-update (ablation) | Co-teaching · CIFAR-10 Symmetry-50% | label precision | Helped |
| Use Joint Training rather than cross-update (ablation) | Co-teaching · CIFAR-10 Symmetry-50% | test accuracy | No clear effect |
| Increase the Co-Regularization weight λ (ablation) | JoCoR · MNIST Symmetry-50% | test accuracy | Helped |
| Increase the Co-Regularization weight λ (ablation) | JoCoR · MNIST Symmetry-50% | label precision | Helped |
Closest papers by meaning
- How does Disagreement Help Generalization against Label Corruption? (ICML 2019 · similarity 0.91)
- Jo-SRC: A Contrastive Approach for Combating Noisy Labels (CVPR 2021 · similarity 0.90)
- Co-teaching: Robust training of deep neural networks with extremely noisy labels (NeurIPS 2018 · similarity 0.90)
- Learning With Neighbor Consistency for Noisy Labels (CVPR 2022 · similarity 0.89)
- Pre-Trained Vision-Language Models as Noisy Partial Annotators (AAAI 2025 · similarity 0.88)
- Learning with Structural Labels for Learning with Noisy Labels (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.