Research map / 3D vision / Human pose estimation
Attention-Aware Compositional Network 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 |
|---|---|---|---|
| Use pose-guided part attention and attention-aware composition instead of global-only features (ablation) | AACN · CUHK03 labeled | R-1: 86.11 -> 90.58 | Helped |
| Use pose-guided part attention and attention-aware composition instead of global-only features (ablation) | AACN · CUHK03 detected | R-1: 83.78 -> 87.98 | Helped |
| Use rigid-part attention only (ablation) | AACN · CUHK03 labeled | R-1: 86.11 -> 87.35 | Helped |
| Use non-rigid-part attention only (ablation) | AACN · CUHK03 labeled | R-1: 86.11 -> 89.89 | Helped |
| Replace PPA localization with RoI localization (ablation) | AFC · CUHK03 detected | R-1: 87.98 -> 86.44 | Hurt |
| Replace PPA localization with parsing localization (ablation) | AFC · CUHK03 detected | R-1: 87.98 -> 82.92 | Hurt |
| Add visibility-score weighting of part features (ablation) | AACN · CUHK03 detected | R-1: 87.98 -> 89.51 | Helped |
| Add visibility-score weighting of part features (ablation) | AACN · Market-1501 SQ | R-1: 86.58 -> 88.69 | Helped |
| Replace GoogleNet with the Spindle base network (ablation) | AACN · CUHK03 labeled | R-1: 91.39 -> 89.16 | Hurt |
| Replace GoogleNet with the Spindle base network (ablation) | AACN · CUHK03 detected | R-1: 89.51 -> 86.65 | Hurt |
| Add the re-ranking method | AACN · Market-1501 Single Query | mAP: 66.87 -> 82.96 | Helped |
Closest papers by meaning
- Learning Deep Context-Aware Features Over Body and Latent Parts for Person Re-Identification (CVPR 2017 · similarity 0.90)
- When Person Re-Identification Meets Event Camera: A Benchmark Dataset and an Attribute-Guided Re-Identification Framework (AAAI 2026 · similarity 0.89)
- From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization (CVPR 2025 · similarity 0.89)
- Exploring the Camera Bias of Person Re-identification (ICLR 2025 · similarity 0.88)
- CLIMB-ReID: A Hybrid CLIP-Mamba Framework for Person Re-Identification (AAAI 2025 · similarity 0.87)
- Instruct-ReID: A Multi-purpose Person Re-identification Task with Instructions (CVPR 2024 · 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.