Research map / Detection and segmentation / Open-set and few-shot detection
C2AM Loss: Chasing a Better Decision Boundary for Long-Tail Object Detection
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 |
|---|---|---|---|
| Replace cross-entropy with C2AM Loss | Mask R-CNN, ResNet-50 · LVIS v1.0 val | AP m: 20.5 -> 25.7 | Helped |
| Replace cross-entropy with C2AM Loss | Mask R-CNN, ResNet-101 · LVIS v1.0 val | AP m: 21.8 -> 27 | Helped |
| Replace cross-entropy on all three heads with C2AM Loss | Cascade Mask R-CNN, ResNet-50 · LVIS v1.0 val | AP m: 22.7 -> 27.6 | Helped |
| Replace cross-entropy on all three heads with C2AM Loss | Cascade Mask R-CNN, ResNet-101 · LVIS v1.0 val | AP m: 24.3 -> 29.2 | Helped |
| Use C2AM Loss instead of the strongest listed baseline | Mask R-CNN, R-50-FPN · LVIS v0.5 | AP m: 28.4 -> 29.7 | Helped |
| Use C2AM Loss instead of the strongest listed baseline | Mask R-CNN, R-50-FPN · LVIS v1.0 | AP m: 26.6 -> 27.2 | Helped |
| Use C2AM Loss instead of the strongest listed baseline | Mask R-CNN, R-101-FPN · LVIS v1.0 | AP m: 28.1 -> 28.6 | Helped |
| Add category-aware angular margin rather than using no margin (ablation) | Mask R-CNN, ResNet-50 · LVIS v1.0 val | AP m: 24.1 -> 25.7 | Helped |
| Replace fixed margin with category-aware adaptive margin (ablation) | Mask R-CNN, ResNet-50 · LVIS v1.0 val | AP m: 24.3 -> 25.7 | Helped |
| Set scale factor s to 10 rather than 30 (ablation) | C2AM Loss on Mask R-CNN, ResNet-50 · LVIS v1.0 val | AP m: 25.7 -> 14.2 | Hurt |
| Set scale factor s to 20 rather than 30 (ablation) | C2AM Loss on Mask R-CNN, ResNet-50 · LVIS v1.0 val | AP m: 25.7 -> 25.4 | Hurt |
| Set scale factor s to 40 rather than 30 (ablation) | C2AM Loss on Mask R-CNN, ResNet-50 · LVIS v1.0 val | AP m: 25.7 -> 24.8 | Hurt |
Closest papers by meaning
- Pursuing Better Decision Boundaries for Long-Tailed Object Detection via Category Information Amount (ICLR 2025 · similarity 0.91)
- Equalized Focal Loss for Dense Long-Tailed Object Detection (CVPR 2022 · similarity 0.89)
- SimLTD: Simple Supervised and Semi-Supervised Long-Tailed Object Detection (CVPR 2025 · similarity 0.86)
- Long-tailed Object Detection Pretraining: Dynamic Rebalancing Contrastive Learning with Dual Reconstruction (NeurIPS 2024 · similarity 0.86)
- Difficulty-aware Balancing Margin Loss for Long-tailed Recognition (AAAI 2025 · similarity 0.85)
- OFBD: Object-Focused Background Debiasing for Long-Tailed Learning (NeurIPS 2026 · similarity 0.84)
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.