Research map / 3D vision / Human pose estimation
Pose-Aware Person Recognition
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 pose-aware weighting instead of average pooling for classifier fusion (ablation) | Pose-Aware Person Recognition · PIPA test | accuracy: 87.78% -> 89.05% | Helped |
| replace pose-aware weighting with max pooling (ablation) | Pose-Aware Person Recognition · PIPA test | accuracy: 89.05% -> 85.51% | Hurt |
| replace pose-aware weighting with elementwise multiplication (ablation) | Pose-Aware Person Recognition · PIPA test | accuracy: 89.05% -> 86.32% | Hurt |
| jointly train head and upper-body regions rather than train them separately (ablation) | Head and upper-body recognition model · PIPA test | accuracy: 84.01 -> 86.27 | Helped |
| combine head, upper-body, and joint-feature classifiers (ablation) | Jointly trained PIPA base model · PIPA test | accuracy: 86.27 -> 86.96 | Helped |
| use both fc6 and fc7 features instead of fc7 alone (ablation) | Head, upper-body, and face representations · PIPA test | accuracy: 76.81 -> 79.54; 72.26 -> 75.19; 66.83 -> 70.40 | Helped |
| use pose-specific models instead of the pose-agnostic base model (ablation) | Pose-Aware Person Recognition · PIPA frontal, semi-left, and semi-right views | accuracy | Helped |
| use pose-specific profile models rather than the base model (ablation) | Pose-Aware Person Recognition · PIPA extreme profile views | accuracy | Hurt |
| use pose-aware classifier-score fusion rather than concatenation (ablation) | Pose-Aware Person Recognition · PIPA test | accuracy: 87.62% -> 89.05% | Helped |
| use joint training of head and upper body rather than separate training (ablation) | Head and upper-body recognition model · Hannah (without tracks) | accuracy: 31.86 -> 32.92 | Helped |
| use joint training of head and upper body rather than separate training (ablation) | Head and upper-body recognition model · Soccer (without tracks) | accuracy: 17.62 -> 18.35 | Helped |
| merge frame predictions by majority voting over tracks (ablation) | Person recognition on Hannah and Soccer · Hannah and Soccer | accuracy: 40.95 -> 44.46; 20.15 -> 24.31 | Helped |
Closest papers by meaning
- Simple Pose: Rethinking and Improving a Bottom-up Approach for Multi-Person Pose Estimation (AAAI 2020 · similarity 0.88)
- DeepCut: Joint Subset Partition and Labeling for Multi Person Pose Estimation (CVPR 2016 · similarity 0.87)
- Unifying Identification and Context Learning for Person Recognition (CVPR 2018 · similarity 0.87)
- PoseTrack21: A Dataset for Person Search, Multi-Object Tracking and Multi-Person Pose Tracking (CVPR 2022 · similarity 0.87)
- A Multi-Level Contextual Model For Person Recognition in Photo Albums (CVPR 2016 · similarity 0.87)
- Optimizing Human Pose Estimation Through Focused Human and Joint Regions (AAAI 2025 · 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.