Research map / Label-efficient and robust learning / Recognition and long-tail
Learning Deep Representation for Imbalanced Classification
Chen Huang, Yining Li, Chen Change Loy, Xiaoou Tang
Check 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 |
|---|---|---|---|
| Use LMLE-kNN rather than ANet | CelebA face attribute classification · CelebA face attributes, average | balanced accuracy (%): 80 -> 84 | Improved |
| Use LMLE-kNN rather than HED | BSDS500 edge detection · BSDS500 | ODS: 0.78 -> 0.78 | No clear effect |
| Add quintuplet sampling and triple-header loss instead of triplet loss (ablation) | Imbalanced face attribute classification · CelebA face attributes | average accuracy (%): 72.43 -> 84.26 | Improved |
| Add quintuplet sampling and triple-header loss instead of triplet loss (ablation) | Imbalanced edge detection · BSDS500 | ODS: 0.74 -> 0.78 | Improved |
| Add resampling to Softmax (ablation) | Face attribute classification · CelebA face attributes | average accuracy (%): 68.07 -> 69.43 | Improved |
| Add cost-sensitive training after resampling to Softmax (ablation) | Face attribute classification · CelebA face attributes | average accuracy (%): 69.43 -> 70.16 | Improved |
| Add resampling to Triplet (ablation) | Face attribute classification · CelebA face attributes | average accuracy (%): 71.29 -> 71.75 | Improved |
| Add cost-sensitive training after resampling to Triplet (ablation) | Face attribute classification · CelebA face attributes | average accuracy (%): 71.75 -> 72.43 | Improved |
| Add resampling to Quintuplet (ablation) | Face attribute classification · CelebA face attributes | average accuracy (%): 81.31 -> 83.39 | Improved |
| Add cost-sensitive training after resampling to Quintuplet (ablation) | Face attribute classification · CelebA face attributes | average accuracy (%): 83.39 -> 84.26 | Improved |
| Add resampling to Quintuplet (ablation) | Edge detection · BSDS500 | ODS: 0.76 -> 0.77 | Improved |
| Add cost-sensitive training after resampling to Quintuplet (ablation) | Edge detection · BSDS500 | ODS: 0.77 -> 0.78 | Improved |
Closest papers by meaning
- Deep Imbalanced Learning for Face Recognition and Attribute Prediction (TPAMI 2019 · similarity 0.92)
- Imbalanced Deep Learning by Minority Class Incremental Rectification (TPAMI 2018 · similarity 0.88)
- Simplifying Neural Network Training Under Class Imbalance (NeurIPS 2023 · similarity 0.88)
- Synthetic Tabular Data Generation for Imbalanced Classification: The Surprising Effectiveness of an Overlap Class (AAAI 2025 · similarity 0.87)
- A Unified Loss for Handling Inter-Class and Intra-Class Imbalance in Medical Image Segmentation (AAAI 2025 · similarity 0.86)
- BaCon: Boosting Imbalanced Semi-supervised Learning via Balanced Feature-Level Contrastive Learning (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.
