Research map / 3D vision / Human pose estimation
A Pose-Sensitive Embedding for Person Re-Identification With Expanded Cross Neighborhood Re-Ranking
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 view information and full-body pose maps to the ResNet-50 embedding (ablation) | ResNet-50 person re-identification · Market-1501 | mAP: 59.8 -> 69.0 | Helped |
| Add view information and full-body pose maps to the ResNet-50 embedding (ablation) | ResNet-50 person re-identification · Duke | mAP: 50.3 -> 62.0 | Helped |
| Add coarse view information (ablation) | ResNet-50 embedding · Market-1501 | mAP: 59.8 -> 66.9 | Helped |
| Add fine-grained full-body pose maps (ablation) | ResNet-50 embedding · Market-1501 | mAP: 59.8 -> 61.6 | Helped |
| Add coarse view information (ablation) | ResNet-50 embedding · Duke | mAP: 50.3 -> 56.7 | Helped |
| Add fine-grained full-body pose maps (ablation) | ResNet-50 embedding · Duke | mAP: 50.3 -> 53.1 | Helped |
| Use both view and pose cues rather than the stronger single cue (ablation) | ResNet-50 embedding · Market-1501 | mAP: 66.9 -> 69.0 | Helped |
| Use both view and pose cues rather than the stronger single cue (ablation) | ResNet-50 embedding · Duke | mAP: 56.7 -> 62.0 | Helped |
| Aggregate expanded cross-neighborhood distances using rank-list distances | IDE-R re-ranking · Market-1501 | mAP: 55.0 -> 71.1 | Helped |
| Aggregate expanded cross-neighborhood distances using original Euclidean distances (ablation) | IDE-R re-ranking · Market-1501 | R-1: 78.9 -> 81.7 | Helped |
| Use rank-list distances instead of original Euclidean distances within ECN (ablation) | IDE-R re-ranking · Market-1501 | mAP: 66.7 -> 71.1 | Helped |
| Use ECN aggregation instead of rank-list distance alone (ablation) | IDE-R re-ranking · Market-1501 | mAP: 66.1 -> 71.1 | Helped |
Closest papers by meaning
- Resource Aware Person Re-Identification Across Multiple Resolutions (CVPR 2018 · similarity 0.90)
- Cheb-GR: Rethinking K-nearest Neighbor Search in Re-ranking for Person Re-identification (CVPR 2025 · similarity 0.90)
- Joint Learning of Single-Image and Cross-Image Representations for Person Re-Identification (CVPR 2016 · similarity 0.89)
- Re-Ranking Person Re-Identification With k-Reciprocal Encoding (CVPR 2017 · similarity 0.89)
- Deep Neural Networks with Inexact Matching for Person Re-Identification (NeurIPS 2016 · similarity 0.89)
- Norm-Aware Embedding for Efficient Person Search (CVPR 2020 · similarity 0.89)
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.