Research map / Label-efficient and robust learning / Recognition and long-tail
BV-Person: A Large-Scale Dataset for Bird-View Person Re-Identification
Cheng Yan, Guansong Pang, Lei Wang, Jile Jiao, Xuetao Feng, Chunhua Shen, Jingjing Li
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 |
|---|---|---|---|
| add Cross Attention layer (ablation) | Baseline · BV-Person-80H | Rank-1: 22.1 -> 28.1 | Improved |
| add Cross Attention layer (ablation) | Baseline · BV-Person-80E | Rank-1: 53.0 -> 59.9 | Improved |
| add Cross Attention layer (ablation) | Baseline · BV-Person-60E | mAP: 50.2 -> 56.2 | Improved |
| add Multi-scale Attention on top of Cross Attention (ablation) | Baseline + CA · BV-Person-60H | Rank-1: 46.3 -> 48.5 | Improved |
| add Multi-scale Attention on top of Cross Attention (ablation) | Baseline + CA · BV-Person-80E | mAP: 45.1 -> 48.6 | Improved |
| add Feature Reconstruction layer (ablation) | Baseline + CA + MA · BV-Person-80H | Rank-1: 30.9 -> 31.8 | Improved |
| add Feature Reconstruction layer (ablation) | Baseline + CA + MA · BV-Person-80E | Rank-1: 62.3 -> 64.4 | Improved |
| add Feature Reconstruction layer (ablation) | Baseline + CA + MA · BV-Person-60E-R | mAP: 65.6 -> 67.0 | Improved |
| use Cross Attention instead of self-attention (ablation) | Baseline · BV-Person-80E | Rank-1: 55.1 -> 59.9 | Improved |
| add self-attention (ablation) | Baseline · BV-Person-80H | Rank-1: 22.1 -> 27.6 | Improved |
| full model vs best baseline (Baseline*) | Baseline* · BV-Person-80E | Rank-1: 55.1 -> 64.4 | Improved |
| full model on general ReID | Baseline* · Market1501 | mAP: 89.2 vs 89.5 (HAA) | No clear effect |
Closest papers by meaning
- Person Re-Identification by Discriminative Selection in Video Ranking (TPAMI 2016 · similarity 0.89)
- Exploring the Camera Bias of Person Re-identification (ICLR 2025 · similarity 0.88)
- Person De-reidentification: A Variation-guided Identity Shift Modeling (CVPR 2025 · similarity 0.88)
- Instruct-ReID: A Multi-purpose Person Re-identification Task with Instructions (CVPR 2024 · similarity 0.87)
- AG-VPReID: A Challenging Large-Scale Benchmark for Aerial-Ground Video-based Person Re-Identification (CVPR 2025 · similarity 0.87)
- From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization (CVPR 2025 · similarity 0.87)
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.
