Research map / Label-efficient and robust learning / Recognition and long-tail
Unsupervised Person Re-Identification with Wireless Positioning under Weak Scene Labeling
Yiheng Liu, Wengang Zhou, Qiaokang Xie, Houqiang 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 pseudo visual label training (second stage) (ablation) | Initial Training · WP-ReID | mAP: 37.9 -> 54.7 | Improved |
| add MMGN with MGM and MMDA wireless similarity (ablation) | Baseline · WP-ReID | mAP: 54.7 -> 66.3 | Improved |
| add MMGN with MGM and MMDA wireless similarity (ablation) | Baseline · WP-ReID | Rank-1: 85.6 -> 91.0 | Improved |
| add MMGN with MGM and MMDA wireless similarity (ablation) | Baseline · Campus4K | mAP: 80.8 -> 86.6 | Improved |
| use MGM histogram-based adjacency instead of plain averaged wireless similarity with GCN (ablation) | BL + MMGN · WP-ReID | mAP: 60.0 -> 66.3 | Improved |
| use MGM adjacency instead of GAT-learned adjacency (ablation) | BL + MMGN · WP-ReID | mAP: 61.3 -> 66.3 | Improved |
| use MGM adjacency instead of GAT-learned adjacency (ablation) | BL + MMGN · Campus4K | mAP: 84.6 -> 86.6 | Improved |
| train MMGN with cross-entropy only instead of adding triplet loss (ablation) | MMGN · WP-ReID | mAP: 58.8 -> 66.3 | Improved |
| train MMGN with cross-entropy only instead of triplet loss (ablation) | MMGN · Campus4K | mAP: 83.4 -> 86.6 | Improved |
| replace baseline with CCUP contrastive baseline and add MMDA and MMGN (ablation) | CCUP · Campus4K | mAP: 88.3 -> 93.7 | Improved |
| replace baseline with CCUP contrastive baseline and add MMDA and MMGN | CCUP · WP-ReID | mAP: 35.7 -> 72.2 | Improved |
| use CCUP baseline instead of original baseline in UMTF | UMTF · WP-ReID | Rank-1: 91.0 -> 84.6 | Hurt |
Closest papers by meaning
- Tracklet Self-Supervised Learning for Unsupervised Person Re-Identification (AAAI 2020 · similarity 0.87)
- Unsupervised Graph Association for Person Re-Identification (ICCV 2019 · similarity 0.87)
- Unsupervised Cross-Dataset Person Re-Identification by Transfer Learning of Spatial-Temporal Patterns (CVPR 2018 · similarity 0.86)
- Unsupervised Person Re-identification by Deep Learning Tracklet Association (ECCV 2018 · similarity 0.86)
- Weakly Supervised Tracklet Association Learning With Video Labels for Person Re-Identification (TPAMI 2023 · similarity 0.86)
- Unsupervised Tracklet Person Re-Identification (TPAMI 2019 · 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.
