Research map / 3D vision / Human pose estimation
Forward Propagation, Backward Regression, and Pose Association for Hand Tracking in the Wild
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 HandLer instead of TraDeS | hand tracking · YouTube-Hand test | MOTA: 52.7 -> 70.0 | Helped |
| use HandLer instead of MPNTrack | hand tracking · VIVA | MOTA: 74.7 -> 87.2 | Helped |
| use HandLer instead of SORT | hand tracking · BSL | MOTA: 63.2 -> 72.5 | Helped |
| use HandLer instead of LightTrack | tracking wrist points using Box2Point · PoseTrack split of YouTube-Hand | MOTA: 60.7 -> 69.6 | Helped |
| use HandLer instead of HandCNN | hand detection · YouTube-Hand | VOC average precision: 72.4 -> 84.1 | Helped |
| use HandLer instead of HandCNN | hand detection · YouTube-Hand Blur&Occ Split | VOC average precision: 62.8(13.1% ↓) -> 76.7(8.8% ↓) | Helped |
| add forward propagation (ablation) | HandLer · YouTube-Hand test | MOTA: 66.1 -> 70.0 | Helped |
| add backward regression (ablation) | HandLer · YouTube-Hand test | MOTA: 65.4 -> 70.0 | Helped |
| add pose association (ablation) | HandLer · YouTube-Hand test | MOTA: 66.1 -> 70.0 | Helped |
| add forward propagation while omitting pose association (ablation) | HandLer-NP · YouTube-Hand test | MOTA: 63.4 -> 66.1 | Helped |
| add backward regression while omitting pose association (ablation) | HandLer-NP · YouTube-Hand test | MOTA: 64.4 -> 66.1 | Helped |
| reduce frame rate from 30 fps to 2 fps (K = 1 to K = 15) | HandLer · YouTube-Hand | MOTA: 70.0 -> 62.7 | Hurt |
Closest papers by meaning
- Whose Hands Are These? Hand Detection and Hand-Body Association in the Wild (CVPR 2022 · similarity 0.89)
- ComPose: When to Trust Hands for Object Pose Tracking (NeurIPS 2026 · similarity 0.88)
- SiMHand: Mining Similar Hands for Large-Scale 3D Hand Pose Pre-training (ICLR 2025 · similarity 0.86)
- FoundHand: Large-Scale Domain-Specific Learning for Controllable Hand Image Generation (CVPR 2025 · similarity 0.86)
- The Surprising Effectiveness of Video Diffusion Models for Hand Motion Reconstruction (NeurIPS 2026 · similarity 0.86)
- StableHand: Quality-Aware Flow Matching for World-Space Dual-Hand Motion Estimation from Egocentric Video (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.