Research map / Label-efficient and robust learning / Recognition and long-tail
Improving Calibration for Long-Tailed Recognition
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 MiSLAS instead of LWS+mixup | ResNet-32 · CIFAR-10-LT IF 100 | top-1 accuracy (%): 79.1 -> 82.1 | Helped |
| Use MiSLAS instead of LWS+mixup | ResNet-32 · CIFAR-100-LT IF 100 | ECE (%): 22.5 -> 4.83 | Helped |
| Add mixup in Stage-1 representation learning (ablation) | cRT on ImageNet-LT · ImageNet-LT | top-1 accuracy (%): 50.3 -> 51.7 | Helped |
| Add mixup in Stage-1 representation learning (ablation) | LWS on ImageNet-LT · ImageNet-LT | top-1 accuracy (%): 51.2 -> 52.0 | Helped |
| Add mixup to Stage-2 as well as Stage-1 (ablation) | cRT on ImageNet-LT · ImageNet-LT | top-1 accuracy (%): 51.7 -> 51.6 | Hurt |
| Add label-aware smoothing instead of class-balanced cross-entropy (ablation) | ResNet-32 classifier training · CIFAR-100-LT IF 100 | ECE (%): 20.2 -> 4.83 | Helped |
| Add label-aware smoothing instead of class-balanced cross-entropy (ablation) | ResNet-32 classifier training · CIFAR-100-LT IF 100 | top-1 accuracy (%): 44.3 -> 47.0 | Helped |
| Update BN running statistics for the Stage-2 sampler (ablation) | LWS · CIFAR-100-LT IF 100 | top-1 accuracy (%): +1.1% | Helped |
| Choose concave rather than the other tested smoothing-function forms (ablation) | LAS · CIFAR-100-LT IF 100 | top-1 accuracy (%) | Helped |
| Vary the smoothing-function form (ablation) | LAS · CIFAR-100-LT IF 100 | top-1 accuracy (%): 0.3% increase | No clear effect |
| Set epsilon_1=0.3 and epsilon_K=0.0 (ablation) | LAS · CIFAR-10-LT IF 100 | top-1 accuracy (%): 3.3% | Helped |
| Set epsilon_1=0.4 and epsilon_K=0.1 (ablation) | LAS · CIFAR-100-LT IF 100 | top-1 accuracy (%): 0.9% | Helped |
Closest papers by meaning
- Long-tailed Recognition with Model Rebalancing (NeurIPS 2025 · similarity 0.89)
- Decoupled Optimisation for Long-Tailed Visual Recognition (AAAI 2024 · similarity 0.88)
- Long-Tailed Recognition via Information-Preservable Two-Stage Learning (NeurIPS 2025 · similarity 0.88)
- Out-of-Distribution Detection in Long-Tailed Recognition with Calibrated Outlier Class Learning (AAAI 2024 · similarity 0.87)
- Difficulty-aware Balancing Margin Loss for Long-tailed Recognition (AAAI 2025 · similarity 0.86)
- Long-Tailed Learning as Multi-Objective Optimization (AAAI 2024 · similarity 0.86)
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.