Research map / Label-efficient and robust learning / Recognition and long-tail
Relation-Aware Global Attention for Person Re-Identification
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 |
|---|---|---|---|
| add RGA-SC to the ResNet-50 baseline | ResNet-50 re-identification model · CUHK03 labeled | mAP: 69.0 -> 77.4 | Helped |
| add RGA-SC to the ResNet-50 baseline | ResNet-50 re-identification model · Market1501 | mAP: 83.7 -> 88.4 | Helped |
| add RGA-SC to the ResNet-50 baseline | ResNet-50 re-identification model · MSMT17 | mAP: 51.5 -> 57.5 | Helped |
| add RGA-S to the baseline (ablation) | ResNet-50 re-identification model · CUHK03 labeled | mAP: 69.0 -> 74.7 | Helped |
| add RGA-C to the baseline (ablation) | ResNet-50 re-identification model · CUHK03 labeled | mAP: 69.0 -> 75.6 | Helped |
| add original feature input to RGA-S (ablation) | RGA-S · CUHK03 labeled | mAP: 72.3 -> 74.7 | Helped |
| add global relation input to RGA-S (ablation) | RGA-S · CUHK03 labeled | mAP: 72.3 -> 74.7 | Helped |
| add original feature input to RGA-C (ablation) | RGA-C · CUHK03 labeled | mAP: 73.7 -> 75.6 | Helped |
| add global relation input to RGA-C (ablation) | RGA-C · CUHK03 labeled | mAP: 73.7 -> 75.6 | Helped |
| combine spatial and channel attention sequentially, spatial first (ablation) | RGA-SC · CUHK03 labeled | mAP: 75.6 -> 77.4 | Helped |
| use spatial and channel attention in parallel instead of sequentially (ablation) | RGA · CUHK03 labeled | mAP: 77.4 -> 73.4 | Hurt |
| reverse the sequence to channel then spatial attention (ablation) | RGA · CUHK03 labeled | mAP: 77.4 -> 75.5 | Hurt |
Closest papers by meaning
- Relation-Guided Spatial Attention and Temporal Refinement for Video-Based Person Re-Identification (AAAI 2020 · similarity 0.91)
- Person Re-Identification Using Heterogeneous Local Graph Attention Networks (CVPR 2021 · similarity 0.90)
- Harmonious Attention Network for Person Re-Identification (CVPR 2018 · similarity 0.90)
- Relation Network for Person Re-Identification (AAAI 2020 · similarity 0.90)
- Diversity Regularized Spatiotemporal Attention for Video-Based Person Re-Identification (CVPR 2018 · similarity 0.90)
- Cheb-GR: Rethinking K-nearest Neighbor Search in Re-ranking for Person Re-identification (CVPR 2025 · 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.