Research map / Video / Video segmentation and tracking
Person Re-Identification by Discriminative Selection in Video Ranking
Taiqing Wang, Shaogang Gong, Xiatian Zhu, Shengjin Wang
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 |
|---|---|---|---|
| use HOG3D + colour (ColHOG3D) instead of HOG3D only (ablation) | DVR(single) · PRID2011 | Rank-1 CMC (%): 28.9 -> 38.9 | Improved |
| use HOG3D + colour (ColHOG3D) instead of HOG3D only (ablation) | DVR(single) · iLIDS-VID | Rank-1 CMC (%): 23.3 -> 36.8 | Improved |
| use HOG3D + colour (ColHOG3D) instead of HOG3D only (ablation) | DVR(single) · HDA+ (2fps) | Rank-1 CMC (%): 18.0 -> 52.0 | Improved |
| select fragments with DVR instead of MS-ColHOG3D averaging (ablation) | DVR · iLIDS-VID | Rank-1 CMC (%): 19.9 -> 39.5 | Improved |
| select fragments with DVR instead of MS-ColHOG3D averaging (ablation) | DVR · PRID2011 | Rank-1 CMC (%): 29.6 -> 40.0 | Improved |
| use single hard selection (DVR(single)) instead of float weighted combination (ablation) | DVR · PRID2011 / iLIDS-VID | Rank-1 CMC (%): 38.9 vs 38.9 | No clear effect |
| select single hard fragment pair instead of float combination (ablation) | DVR · iLIDS-VID selecting time | Rank-1 CMC (%): 740 -> 97 seconds | Improved |
| select top 3 fragment pairs instead of single pair (ablation) | DVR · iLIDS-VID | Rank-1 CMC (%): 36.8 -> 39.5 | Improved |
| select top 5 fragment pairs instead of top 3 (ablation) | DVR · PRID2011 | Rank-1 CMC (%): 40.0 -> 40.8 | Improved |
| select top 5 fragment pairs instead of top 3 (training time) (ablation) | DVR · iLIDS-VID ranking time | Rank-1 CMC (%): 58 -> 81 seconds | Hurt |
| add DVR score to MS-ColLBP | MS-ColLBP · iLIDS-VID | Rank-1 CMC (%): 23.2 -> 41.0 | Improved |
| add DVR score to eSDC | eSDC · iLIDS-VID | Rank-1 CMC (%): 10.2 -> 41.0 | Improved |
Closest papers by meaning
- When Person Re-Identification Meets Event Camera: A Benchmark Dataset and an Attribute-Guided Re-Identification Framework (AAAI 2026 · similarity 0.90)
- Spatial-Temporal Person Re-Identification (AAAI 2019 · similarity 0.89)
- BV-Person: A Large-Scale Dataset for Bird-View Person Re-Identification (ICCV 2021 · similarity 0.89)
- Exploring the Camera Bias of Person Re-identification (ICLR 2025 · similarity 0.88)
- From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization (CVPR 2025 · similarity 0.88)
- Instruct-ReID: A Multi-purpose Person Re-identification Task with Instructions (CVPR 2024 · similarity 0.88)
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.
