Research map / Label-efficient and robust learning / Noisy labels
Learning from Noisy Labels with Complementary Loss Functions
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 complementary CE+MAE with mixup instead of DivideMix | Noisy-label classification with PreAct-ResNet-18 · CIFAR-10, uniform noise 0.4 | accuracy (%): 93.62±0.15 -> 94.79±0.11 | Helped |
| Use complementary CE+MAE with mixup instead of DivideMix | Noisy-label classification with PreAct-ResNet-18 · CIFAR-100, uniform noise 0.6 | accuracy (%): 69.38±0.26 -> 72.17±0.30 | Helped |
| Use complementary CE+MAE with mixup instead of DivideMix | Noisy-label classification with PreAct-ResNet-18 · CIFAR-100, class-conditional noise 0.4 | accuracy (%): 50.79±0.62 -> 68.07±0.79 | Helped |
| Add robust MAE loss to the self-ensembled CE method (ablation) | Ours* (CE) on CIFAR-100 · CIFAR-100, uniform noise 0.6 | accuracy (%): 65.09±0.33 -> 72.17±0.30 | Helped |
| Add robust MAE loss to the self-ensembled CE method (ablation) | Ours* (CE) on CIFAR-10 · CIFAR-10, uniform noise 0.6 | accuracy (%): 92.33±0.24 -> 93.59±0.19 | Helped |
| Replace MAE with APL in the mixup complementary-loss method (ablation) | Ours* on CIFAR-100 · CIFAR-100, class-conditional noise 0.4 | accuracy (%): 68.07±0.79 vs 68.18±1.16 | Mixed |
| Apply complementary loss to real-world instance-dependent noisy labels | Clothing1M classifier · Clothing1M | accuracy (%): 74.76 -> 73.59 | Hurt |
Closest papers by meaning
- Normalized Loss Functions for Deep Learning with Noisy Labels (ICML 2020 · similarity 0.92)
- On Learning Contrastive Representations for Learning With Noisy Labels (CVPR 2022 · similarity 0.91)
- Robust Loss Functions under Label Noise for Deep Neural Networks (AAAI 2017 · similarity 0.91)
- Asymmetric Loss Functions for Learning with Noisy Labels (ICML 2021 · similarity 0.90)
- $\epsilon$-Softmax: Approximating One-Hot Vectors for Mitigating Label Noise (NeurIPS 2024 · similarity 0.88)
- Enhancing Noise-Robust Losses for Large-Scale Noisy Data Learning (AAAI 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.