Research map / Detection and segmentation / CNN features
Pyramidal Person Re-IDentification via Multi-Loss Dynamic Training
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 |
|---|---|---|---|
| use the coarse-to-fine pyramid with local, global, and intermediate-scale branches | Pyramid-ours · Market-1501 | mAP: 88.2 vs 86.9 | Helped |
| use the coarse-to-fine pyramid with local, global, and intermediate-scale branches | Pyramid-ours · DukeMTMC-reID | mAP: 79.0 vs 78.4 | Helped |
| use the coarse-to-fine pyramid with local, global, and intermediate-scale branches | Pyramid-ours · CUHK03 labelled | mAP: 76.9 vs 67.4 | Helped |
| use the coarse-to-fine pyramid with local, global, and intermediate-scale branches | Pyramid-ours · CUHK03 detected | mAP: 74.8 vs 66.0 | Helped |
| add lower-level local branches instead of using only the global branch (ablation) | Pyramid-ours · Market-1501 | mAP: 82.1 -> 84.9 | Helped |
| add more pyramid branches (ablation) | Pyramid-ours · Market-1501 | mAP: 82.1; 84.9; 86.7; 87.2; 87.5; 88.2 | Helped |
| reduce branch feature dimension to 64 (ablation) | Pyramid-ours · Market-1501 | mAP: 88.2 -> 86.9 | Hurt |
| increase branch feature dimension to 256 (ablation) | Pyramid-ours · Market-1501 | mAP: 88.2 -> 87.8 | Hurt |
| remove triplet loss and train with identification loss only (ablation) | Pyramid-ours · Market-1501 | mAP: 88.2 -> 86.5 | Hurt |
Closest papers by meaning
- Horizontal Pyramid Matching for Person Re-Identification (AAAI 2019 · similarity 0.90)
- From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization (CVPR 2025 · similarity 0.88)
- A Bottom-Up Clustering Approach to Unsupervised Person Re-Identification (AAAI 2019 · similarity 0.87)
- Multi-Channel Pyramid Person Matching Network for Person Re-Identification (AAAI 2018 · similarity 0.87)
- DC-Former: Diverse and Compact Transformer for Person Re-identification (AAAI 2023 · similarity 0.87)
- Person30K: A Dual-Meta Generalization Network for Person Re-Identification (CVPR 2021 · 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.