Research map / Label-efficient and robust learning / Recognition and long-tail
COCAS: A Large-Scale Clothes Changing Person Dataset for Re-Identification
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 |
|---|---|---|---|
| Add the clothes feature branch to the biometric feature branch (ablation) | BC-Net · COCAS test | mAP: 12.2 -> 46.8 | Helped |
| Add the biometric feature branch to the clothes feature branch (ablation) | BC-Net · COCAS test | mAP: 28.7 -> 46.8 | Helped |
| Add the biometric-feature identification loss (ablation) | BC-Net · COCAS test | mAP: 32.7 -> 46.8 | Helped |
| Add triplet loss (ablation) | BC-Net · COCAS test | mAP: 42.8 -> 46.8 | Helped |
| Add the biometric-feature mask module (ablation) | BC-Net · COCAS test | mAP: 43.6 -> 46.8 | Helped |
| Add the clothes detector for target images (ablation) | BC-Net · COCAS test | mAP: 39.5 -> 46.8 | Helped |
| Use raw rather than desensitized images (ablation) | BC-Net · COCAS test | mAP: 46.8 -> 52.8; 54.8 -> 63.7; 57.2 -> 65.1; 68.5 -> 75.4 | Helped |
| Use XQDA instead of Euclidean distance | BC-Net (desensitized) · COCAS test | mAP: 46.8 -> 57.2 | Helped |
| Apply re-ranking with Euclidean distance | BC-Net (desensitized) · COCAS test | mAP: 46.8 -> 54.8 | Helped |
| Combine XQDA with re-ranking | BC-Net (desensitized) · COCAS test | mAP: 46.8 -> 68.5 | Helped |
| Train using provided clothes templates for queries | BC-Net · COCAS test | mAP: +37.6% mAP | Helped |
Closest papers by meaning
- Identity-Clothing Similarity Modeling for Unsupervised Clothing Change Person Re-Identification (CVPR 2025 · similarity 0.89)
- DIFFER: Disentangling Identity Features via Semantic Cues for Clothes-Changing Person Re-ID (CVPR 2025 · similarity 0.89)
- Clothes-Changing Person Re-Identification With RGB Modality Only (CVPR 2022 · similarity 0.88)
- Graph-Enhanced Attribute-Aware Modeling for Cloth-Changing Person Re-Identification (NeurIPS 2026 · similarity 0.87)
- DKC: Differentiated Knowledge Consolidation for Cloth-Hybrid Lifelong Person Re-identification (CVPR 2025 · similarity 0.85)
- Instruct-ReID: A Multi-purpose Person Re-identification Task with Instructions (CVPR 2024 · 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.