Research map / 3D vision / Human pose estimation
Seeing Invisible Poses: Estimating 3D Body Pose From Egocentric Video
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 the full pose-path method rather than AlwaysStanding | Egocentric 3D pose estimation · Same subject, different environments | NAvgAll (cm): 24.6 -> 17.2 | Helped |
| Use the full pose-path method rather than AlwaysStanding | Egocentric 3D pose estimation · Different subjects and environments | NAvgAll (cm): 22.4 -> 19.9 | Helped |
| Add unrestricted pose-level selection and refinement instead of restricting predictions to first-stage cluster centers (ablation) | Pose-path method · Same subject, different environments | NAvgAll (cm): 19.1 -> 17.2 | Helped |
| Add unrestricted pose-level selection and refinement instead of restricting predictions to first-stage cluster centers (ablation) | Pose-path method · Different subjects and environments | NAvgAll (cm): 24.6 -> 19.9 | Helped |
| Use handcrafted homography features instead of CNN features trained from scratch (ablation) | Pose-path method · Same subject, different environments | NAvgAll (cm): 48.1 -> 17.2 | Helped |
| Use handcrafted homography features instead of fine-tuned CNN features (ablation) | Pose-path method · Same subject, different environments | NAvgAll (cm): 48.7 -> 17.2 | Helped |
| Use handcrafted homography features instead of CNN features trained from scratch (ablation) | Pose-path method · Different subjects and environments | NAvgAll (cm): 35.4 -> 19.9 | Helped |
| Use handcrafted homography features instead of fine-tuned CNN features (ablation) | Pose-path method · Different subjects and environments | NAvgAll (cm): 36.4 -> 19.9 | Helped |
| Use the full pose-path method rather than CNN-Regression | Egocentric 3D pose estimation · Same subject, different environments | NAvgAll (cm): 29.7 -> 17.2 | Helped |
| Use the full pose-path method rather than KdTree nearest neighbors | Egocentric 3D pose estimation · Different subjects and environments | NAvgAll (cm): 44.5 -> 19.9 | Helped |
| Use the full pose-path method rather than AlwaysStanding | Egocentric 3D pose estimation · Same subject, different environments | NAvg(W+A) (cm): 32.4 -> 19.9 | Helped |
| Use the full pose-path method rather than AlwaysStanding | Egocentric 3D pose estimation · Different subjects and environments | NAvg(W+A) (cm): 28.9 -> 23.6 | Helped |
Closest papers by meaning
- Estimating Ego-Body Pose from Doubly Sparse Egocentric Video Data (NeurIPS 2024 · similarity 0.92)
- EgoSim: An Egocentric Multi-view Simulator and Real Dataset for Body-worn Cameras during Motion and Activity (NeurIPS 2024 · similarity 0.91)
- 3D Human Pose Perception from Egocentric Stereo Videos (CVPR 2024 · similarity 0.89)
- FRAME: Floor-aligned Representation for Avatar Motion from Egocentric Video (CVPR 2025 · similarity 0.89)
- Estimating Body and Hand Motion in an Ego-sensed World (CVPR 2025 · similarity 0.88)
- HOT3D: Hand and Object Tracking in 3D from Egocentric Multi-View Videos (CVPR 2025 · 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.