Research map / Graph learning / Knowledge and temporal graphs
Learning Context Graph for Person Search
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 relative attention to weight matched body-part pairs instead of using uniform weights (ablation) | Person search feature matching · CUHK-SYSU | top-1 (%): 79.2% → 82.7% | Helped |
| Add relative attention to weight matched body-part pairs instead of using uniform weights (ablation) | Person search feature matching · PRW | top-1 (%): 65.4% → 67.8% | Helped |
| Add top-K context pairs and a contextual graph to aggregate context (ablation) | Person search model · CUHK-SYSU | top-1 (%): 82.7 → 86.5 | Helped |
| Add top-K context pairs and a contextual graph to aggregate context (ablation) | Person search model · PRW | top-1 (%): 67.8 → 73.6 | Helped |
| Use three context neighbors rather than adding more neighbors (ablation) | Contextual graph · Person search benchmarks | retrieval performance | Mixed |
| Represent target and context matches as paired nodes in one graph instead of separate Siamese graphs (ablation) | Context graph learning · CUHK-SYSU | top-1: 83.4 → 86.5 | Helped |
| Represent target and context matches as paired nodes in one graph instead of separate Siamese graphs (ablation) | Context graph learning · PRW | mAP: 29.5 → 33.4 | Helped |
| Combine global and body-part features uniformly rather than using only the whole-body feature (ablation) | Person search feature matching · CUHK-SYSU | top-1 (%): 77.5% → 79.2% | Helped |
| Use lower-body features alone rather than upper-body features alone (ablation) | Person search feature matching · CUHK-SYSU | top-1 (%): 65.7 → 61.8 | Hurt |
Closest papers by meaning
- Context-Aware Graph Convolution Network for Target Re-identification (AAAI 2021 · similarity 0.90)
- Unifying Identification and Context Learning for Person Recognition (CVPR 2018 · similarity 0.89)
- Joint Detection and Identification Feature Learning for Person Search (CVPR 2017 · similarity 0.88)
- Sequential End-to-end Network for Efficient Person Search (AAAI 2021 · similarity 0.88)
- Cheb-GR: Rethinking K-nearest Neighbor Search in Re-ranking for Person Re-identification (CVPR 2025 · similarity 0.87)
- Divide and Conquer: Hybrid Pre-training for Person Search (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.