Research map / Label-efficient and robust learning / Recognition and long-tail
Dynamic Curriculum Learning for Imbalanced Data Classification
Yiru Wang, Weihao Gan, Jie Yang, Wei Wu, Junjie Yan
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 sampling scheduler (convex) (ablation) | DeepID2 · CelebA (40 attributes) | mA: 81.17 -> 86.58 | Improved |
| add triplet loss with easy anchors (ablation) | DeepID2 + SS · CelebA (40 attributes) | mA: 86.58 -> 87.55 | Improved |
| add loss scheduler (ablation) | DeepID2 + SS + TL · CelebA (40 attributes) | mA: 87.55 -> 89.05 | Improved |
| use convex instead of linear sampling scheduler (ablation) | DeepID2 + SS · CelebA (40 attributes) | mA: 86.36 -> 86.58 | Improved |
| use convex instead of concave sampling scheduler (ablation) | DeepID2 + SS · CelebA (40 attributes) | mA: 85.90 -> 86.58 | Improved |
| use convex instead of composite sampling scheduler (ablation) | DeepID2 + SS · CelebA (40 attributes) | mA: 86.07 -> 86.58 | Improved |
| power-transform target distribution instead of linear decrease (ablation) | DeepID2 + SS · CelebA (40 attributes) | mA: 85.11 -> 86.58 | Improved |
| add DCL (RAP, ratio 1:1 to 1:25) (ablation) | ResNet-50 baseline · RAP | mA: 79.3 -> 83.1 | Improved |
| add DCL (RAP, ratio 1:25 to 1:50) (ablation) | ResNet-50 baseline · RAP | mA: 68.9 -> 83.9 | Improved |
| add DCL (RAP, ratio above 1:50) (ablation) | ResNet-50 baseline · RAP | mA: 68.0 -> 85.5 | Improved |
| add DCL on balanced data | Cross Entropy · CIFAR-100 | accuracy: 68.1 -> 71.5 | Improved |
| replace CRL with DCL on balanced data | CRL · CIFAR-100 | accuracy: 69.3 -> 71.5 | Improved |
Closest papers by meaning
- Deep Imbalanced Attribute Classification using Visual Attention Aggregation (ECCV 2018 · similarity 0.86)
- Class Rectification Hard Mining for Imbalanced Deep Learning (ICCV 2017 · similarity 0.86)
- On the Pitfalls of Instance-Based Dynamic Curricula (NeurIPS 2026 · similarity 0.85)
- Class-Attribute Priors: Adapting Optimization to Heterogeneity and Fairness Objective (AAAI 2024 · similarity 0.83)
- Gradient Reweighting: Towards Imbalanced Class-Incremental Learning (CVPR 2024 · similarity 0.83)
- Addressing Imbalanced Domain-Incremental Learning through Dual-Balance Collaborative Experts (ICML 2025 · similarity 0.82)
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.
