Research map / Label-efficient and robust learning / Recognition and long-tail
Breadcrumbs: Adversarial Class-Balanced Sampling for Long-Tailed Recognition
Bo Liu, Haoxiang Li, Hao Kang, Gang Hua, Nuno Vasconcelos
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 feature back-tracking without alignment (ablation) | Decoupling (cRT) · ImageNet-LT few-shot, ResNet-10 | accuracy: 21.5 -> 23.8 | Improved |
| add feature back-tracking without alignment (ablation) | Decoupling (cRT) · ImageNet-LT many-shot, ResNet-10 | accuracy: 51.8 -> 50.4 | Hurt |
| add class-specific back-tracking lengths without alignment (ablation) | Decoupling (cRT) · ImageNet-LT many-shot, ResNet-10 | accuracy: 51.8 -> 50.8 | No clear effect |
| add class alignment (Weak Breadcrumb) (ablation) | Decoupling (cRT) · ImageNet-LT overall, ResNet-10 | accuracy: 41.4 -> 43.2 | Improved |
| add class alignment (Weak Breadcrumb) (ablation) | Decoupling (cRT) · ImageNet-LT few-shot, ResNet-10 | accuracy: 21.5 -> 25.1 | Improved |
| use snapshots from throughout stage 1 (Strong vs Weak Breadcrumb) (ablation) | Breadcrumb · ImageNet-LT overall, ResNet-10 | accuracy: 43.2 -> 44.0 | Improved |
| use snapshots from throughout stage 1 (Strong vs Weak Breadcrumb) (ablation) | Breadcrumb · ImageNet-LT few-shot, ResNet-10 | accuracy: 25.1 -> 26.4 | Improved |
| add variance alignment on top of mean alignment (ablation) | Breadcrumb · ImageNet-LT overall, ResNet-10 | accuracy: 44.0 -> 43.9 | No clear effect |
| use class-specific mean alignment instead of class-agnostic alignment (ablation) | Breadcrumb · ImageNet-LT overall, ResNet-10 | accuracy: 38.5 -> 44.0 | Improved |
| replace Decoupling (cRT) with Breadcrumb | cRT · ImageNet-LT few-shot, ResNet-10 | accuracy: 21.5 -> 26.4 | Improved |
| replace Decoupling (cRT) with Breadcrumb | cRT · Places-LT overall, ResNet-152 | accuracy: 37.9 -> 39.3 | Improved |
| replace Decoupling (cRT) with Breadcrumb | cRT · Places-LT many-shot, ResNet-152 | accuracy: 37.8 -> 40.6 | Improved |
Closest papers by meaning
- Decoupling Representation and Classifier for Long-Tailed Recognition (ICLR 2020 · similarity 0.88)
- Deep Long-Tailed Learning: A Survey (TPAMI 2023 · similarity 0.88)
- Difficulty-aware Balancing Margin Loss for Long-tailed Recognition (AAAI 2025 · similarity 0.88)
- Long-tailed Recognition with Model Rebalancing (NeurIPS 2025 · similarity 0.87)
- Exploring Weight Balancing on Long-Tailed Recognition Problem (ICLR 2024 · similarity 0.87)
- Long-Tailed Learning as Multi-Objective Optimization (AAAI 2024 · similarity 0.85)
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.
