Research map / Detection and segmentation / Weakly supervised detection and segmentation
Exploring Visual Context for Weakly Supervised Person Search
Read the paperCheck 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 detection context to the baseline (ablation) | CGPS · CUHK-SYSU | mAP: 71.2 -> 75.5 (+4.3) | Helped |
| add memory context alone to the baseline (ablation) | CGPS · CUHK-SYSU | mAP: 71.2 -> 71.3 (+0.1) | Helped |
| add scene context to the baseline (ablation) | CGPS · CUHK-SYSU | mAP: 71.2 -> 77.1 (+5.9) | Helped |
| combine detection, memory, and scene context (ablation) | CGPS · CUHK-SYSU | mAP: 71.2 -> 80.0 (+8.8) | Helped |
| use cumulative-confidence hard-negative selection instead of selecting a fixed number of negatives (ablation) | CGPS memory-context loss · CUHK-SYSU | mAP | Helped |
| add the instance-to-instance detection-context term (ablation) | CGPS detection-context loss · CUHK-SYSU | mAP | Helped |
| add the instance-to-background detection-context term (ablation) | CGPS detection-context loss · CUHK-SYSU | mAP | Helped |
| use ResNet-101 instead of ResNet-50 (ablation) | CGPS · CUHK-SYSU | mAP: 80.0 -> 80.4 | Helped |
| add PRW training data to CUHK-SYSU training (ablation) | CGPS · CUHK-SYSU | mAP: 80.0 -> 80.5 | Helped |
| add PRW training data to CUHK-SYSU training (ablation) | CGPS · PRW | mAP: 15.0 -> 18.7 | Helped |
| add COCO training data to CUHK-SYSU training (ablation) | CGPS · CUHK-SYSU | mAP: 80.0 -> 74.8 | Hurt |
| replace ground-truth boxes with detector-predicted boxes (ablation) | CGPS · CUHK-SYSU and PRW | mAP: 80.0 -> 76.2; 16.2 -> 14.7 | Hurt |
Closest papers by meaning
- Optimizing Label Assignment for Weakly Supervised Person Search (AAAI 2025 · similarity 0.90)
- Joint Detection and Identification Feature Learning for Person Search (CVPR 2017 · similarity 0.89)
- Diverse Knowledge Distillation for End-to-End Person Search (AAAI 2021 · similarity 0.88)
- Hierarchical Online Instance Matching for Person Search (AAAI 2020 · similarity 0.87)
- Divide and Conquer: Hybrid Pre-training for Person Search (AAAI 2024 · similarity 0.86)
- Cheb-GR: Rethinking K-nearest Neighbor Search in Re-ranking for Person Re-identification (CVPR 2025 · 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.