Research map / 3D vision / Human pose estimation
Recurrent Attention Models for Depth-Based Person 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 |
|---|---|---|---|
| Use 3D RAM instead of Skeleton (NN) | BIWI single-shot identification · BIWI | Top-1 recognition rate (%): 21.1 -> 30.1 | Helped |
| Use 3D RAM instead of 2D RAM (ablation) | IAS-A single-shot identification · IAS-A | Top-1 recognition rate (%): 46.9 -> 48.3 | Helped |
| Use 3D RAM instead of 2D RAM (ablation) | IAS-B single-shot identification · IAS-B | Top-1 recognition rate (%): 61.0 -> 63.7 | Helped |
| Use 3D RAM instead of Skeleton (SVM) | PAVIS single-shot identification · PAVIS | Top-1 recognition rate (%): 35.7 -> 41.3 | Helped |
| Use 3D RAM instead of 2D RAM (ablation) | DPI-T single-shot identification · DPI-T | Top-1 recognition rate (%): 33.8 -> 47.5 | Helped |
| Use 4D RAM instead of the strongest listed depth baseline | BIWI multi-shot identification · BIWI | Top-1 recognition rate (%): 39.3 -> 45.3 | Helped |
| Use 4D RAM instead of 3D CNN+Avg Pooling | IAS-A multi-shot identification · IAS-A | Top-1 recognition rate (%): 33.4 -> 53.5 | Helped |
| Use 4D RAM instead of 3D CNN+Avg Pooling | IAS-B multi-shot identification · IAS-B | Top-1 recognition rate (%): 39.1 -> 64.4 | Helped |
| Use 4D RAM instead of 3D CNN+Avg Pooling | PAVIS multi-shot identification · PAVIS | Top-1 recognition rate (%): 27.5 -> 43.0 | Helped |
| Use 4D RAM instead of 3D CNN+Avg Pooling | DPI-T multi-shot identification · DPI-T | Top-1 recognition rate (%): 28.4 -> 55.6 | Helped |
Closest papers by meaning
- Diversity Regularized Spatiotemporal Attention for Video-Based Person Re-Identification (CVPR 2018 · similarity 0.88)
- Unifying Identification and Context Learning for Person Recognition (CVPR 2018 · similarity 0.87)
- Relation-Aware Global Attention for Person Re-Identification (CVPR 2020 · similarity 0.87)
- SapiensID: Foundation for Human Recognition (CVPR 2025 · similarity 0.86)
- From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization (CVPR 2025 · similarity 0.85)
- SapiensID 2.0: Aligning Human Recognition Foundation Models with Human Perception (NeurIPS 2026 · similarity 0.85)
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.