Image Classification On Clothing1M Using
评估指标
Accuracy
评测结果
各个模型在此基准测试上的表现结果
| Paper Title | Repository | ||
|---|---|---|---|
| CurriculumNet | 81.5% | CurriculumNet: Weakly Supervised Learning from Large-Scale Web Images | |
| Forward | 80.27 | Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approach | |
| CleanNet w_soft | 79.90 | CleanNet: Transfer Learning for Scalable Image Classifier Training with Label Noise | |
| EMLC (k=1) | 79.35% | Enhanced Meta Label Correction for Coping with Label Corruption | |
| DMLP-DivideMix | 78.23% | Learning from Noisy Labels with Decoupled Meta Label Purifier | |
| FasTEN | 77.83% | Learning with Noisy Labels by Efficient Transition Matrix Estimation to Combat Label Miscorrection | |
| PUDistill | 77.70 | Training Classifiers that are Universally Robust to All Label Noise Levels | |
| L2B (ResNet-18) | 77.5 ± 0.2% | L2B: Learning to Bootstrap Robust Models for Combating Label Noise | |
| MLC | 75.78% | Meta Label Correction for Noisy Label Learning |
0 of 9 row(s) selected.