Research map / Label-efficient and robust learning / Noisy labels
NLNL: Negative Learning for Noisy Labels
Youngdong Kim, Junho Yim, Juseung Yun, Junmo Kim
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 SelPL after NL and SelNL (ablation) | SelNLPL · CIFAR10 30% symm-inc | accuracy (%): 92.44 -> 93.82 | Improved |
| add SelPL after NL and SelNL (ablation) | SelNLPL · CIFAR10 50% symm-inc | accuracy (%): 89.36 -> 91.17 | Improved |
| add SelNL between NL and SelPL (ablation) | SelNLPL · CIFAR10 30% symm-inc | accuracy (%): 93.41 -> 93.82 | Improved |
| add SelNL between NL and SelPL (ablation) | SelNLPL · CIFAR10 50% symm-inc | accuracy (%): 72.91 -> 91.17 | Improved |
| add SelNL and SelPL after NL (ablation) | SelNLPL · CIFAR10 30% symm-inc | accuracy (%): 87.32 -> 93.82 | Improved |
| add SelNL and SelPL after NL (filtering precision) (ablation) | SelNLPL · CIFAR10 50% symm-inc | precision: 50.38 -> 95.38 | Improved |
| train with NL (complementary labels) instead of PL | NL · CIFAR10 30% symm-inc | test accuracy | Improved |
| provide 110 random complementary labels per image instead of one (ablation) | NL · CIFAR100 | convergence | Improved |
| use proposed method instead of Truncated Lq | SelNLPL+pseudo labeling · CIFAR10 20% symm-exc | accuracy (%): 89.70 -> 94.23 | Improved |
| use proposed method instead of Joint | SelNLPL+pseudo labeling · CIFAR10 40% asymm (Pre-ResNet32) | accuracy (%): 91.7 -> 90.99 | Hurt |
| use proposed method instead of Semi | SelNLPL+pseudo labeling · CIFAR10 60% asymm (ResNet14) | accuracy (%): 75.8 -> 74.44 | Hurt |
| use proposed method at 80% symm-exc noise | SelNLPL+pseudo labeling · FashionMNIST/CIFAR 80% symm-exc | convergence | Hurt |
Closest papers by meaning
- Joint Negative and Positive Learning for Noisy Labels (CVPR 2021 · similarity 0.95)
- BadLabel: A Robust Perspective on Evaluating and Enhancing Label-Noise Learning (TPAMI 2024 · similarity 0.89)
- Toward Robustness against Label Noise in Training Deep Discriminative Neural Networks (NeurIPS 2017 · similarity 0.88)
- Coresets for Robust Training of Deep Neural Networks against Noisy Labels (NeurIPS 2020 · similarity 0.88)
- Adaptive Integration of Partial Label Learning and Negative Learning for Enhanced Noisy Label Learning (AAAI 2024 · similarity 0.87)
- Learning to Learn From Noisy Labeled Data (CVPR 2019 · 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.
