Research map / 3D vision / Human pose estimation
Personalizing Human Video Pose Estimation
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 spatial matching to spread initial pose annotations (ablation) | Personalized ConvNet pipeline · YouTube Pose Subset | accuracy (%) at d=20 pixels, average: 75.8 -> 83.3 | Helped |
| Add temporal propagation using dense optical flow (ablation) | Personalized ConvNet pipeline · YouTube Pose Subset | accuracy (%) at d=20 pixels, average: 83.3 -> 87.2 | Helped |
| Fine-tune the generic ConvNet using the personalized annotations (ablation) | Pfister et al. ConvNet · YouTube Pose Subset | accuracy (%) at d=20 pixels, average: 87.2 -> 91.0 | Helped |
| Personalize the generic ConvNet using video annotations | Pfister et al. ConvNet · YouTube Pose | accuracy (%) at d=20 pixels, average: 76.9 -> 89.9 | Helped |
| Personalize the generic ConvNet using video annotations | Pfister et al. ConvNet · MPII Cooking | accuracy (%) at d=20 pixels, average: 66.6 -> 81.7 | Helped |
| Personalize the generic ConvNet using video annotations | Pfister et al. ConvNet · BBC Pose | accuracy (%) at d=6 pixels, average: 85.3 -> 95.6 | Helped |
| Personalize the generic ConvNet using video annotations | Pfister et al. ConvNet · UYDP | elbow accuracy (%) at APK=0.2: 35.2 -> 57.6 | Helped |
| Add automatically annotated YouTube frames to training | FLIC-trained generic ConvNet · FLIC test | wrist accuracy at normalized distance 0.1: +4% | Helped |
| Add automatically annotated YouTube frames to training | FLIC-trained generic ConvNet · YouTube Pose | wrist accuracy at d=20: +8% | Helped |
| Add automatically annotated YouTube frames to training | FLIC-trained generic ConvNet · MPII Cooking | elbow accuracy at d=20: +8% | Helped |
Closest papers by meaning
- End-to-End Multi-Person Pose Estimation with Pose-Aware Video Transformer (AAAI 2026 · similarity 0.89)
- Learning Temporal Pose Estimation from Sparsely-Labeled Videos (NeurIPS 2019 · similarity 0.88)
- Harvesting Multiple Views for Marker-Less 3D Human Pose Annotations (CVPR 2017 · similarity 0.88)
- Temporal Feature Alignment and Mutual Information Maximization for Video-Based Human Pose Estimation (CVPR 2022 · similarity 0.88)
- PoseTrack: A Benchmark for Human Pose Estimation and Tracking (CVPR 2018 · similarity 0.87)
- Test-Time Personalization with a Transformer for Human Pose Estimation (NeurIPS 2021 · 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.