Research map / Label-efficient and robust learning / Recognition and long-tail
Horizontal Pyramid Matching for Person Re-Identification
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 a multi-scale horizontal pyramid instead of using only same-sized partitions (ablation) | HPM · Market-1501 | mAP: 71.2 -> 82.7 | Helped |
| increase average-pooled pyramid scales from 1 to 4 (ablation) | HPM · Market-1501 | mAP: 71.2 -> 79.5 | Helped |
| increase average-pooled pyramid scales from 1 to 4 (ablation) | HPM · DukeMTMC-ReID | mAP: 61.0 -> 72.1 | Helped |
| increase average-pooled pyramid scales from 1 to 4 (ablation) | HPM · CUHK03 | mAP: 37.3 -> 53.4 | Helped |
| use max pooling instead of average pooling alone (ablation) | HPM with 4 scales · Market-1501 | mAP: 79.5 -> 81.6 | Helped |
| use max pooling instead of average pooling alone (ablation) | HPM with 4 scales · DukeMTMC-ReID | mAP: 72.1 -> 74.1 | Helped |
| use max pooling instead of average pooling alone (ablation) | HPM with 4 scales · CUHK03 | mAP: 53.4 -> 57.4 | Helped |
| combine average and max pooling rather than using max pooling alone (ablation) | HPM with 4 scales · Market-1501 | mAP: 81.6 -> 82.7 | Helped |
| combine average and max pooling rather than using max pooling alone (ablation) | HPM with 4 scales · DukeMTMC-ReID | mAP: 74.1 -> 74.3 | Helped |
| combine average and max pooling rather than using max pooling alone (ablation) | HPM with 4 scales · CUHK03 | mAP: 57.4 -> 57.5 | Helped |
| add 12 or 24 pyramid scales (ablation) | HPM · person re-identification benchmarks | mAP | No clear effect |
Closest papers by meaning
- Pyramidal Person Re-IDentification via Multi-Loss Dynamic Training (CVPR 2019 · similarity 0.90)
- Relation Network for Person Re-Identification (AAAI 2020 · similarity 0.89)
- Multi-Channel Pyramid Person Matching Network for Person Re-Identification (AAAI 2018 · similarity 0.88)
- Deep Neural Networks with Inexact Matching for Person Re-Identification (NeurIPS 2016 · similarity 0.88)
- From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization (CVPR 2025 · similarity 0.87)
- Instruct-ReID: A Multi-purpose Person Re-identification Task with Instructions (CVPR 2024 · 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.