Research map / 3D vision / Human pose estimation
HOnnotate: A Method for 3D Annotation of Hand and Object Poses
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 |
|---|---|---|---|
| optimize hand and object poses jointly across frames instead of ending with single-frame optimization (ablation) | HOnnotate multi-camera annotation · HO-3D multi-camera annotation | mean hand-joint error (cm): 0.92 (±0.34) -> 0.77 (±0.29) | Helped |
| combine silhouette and depth terms rather than use either alone (ablation) | HOnnotate multi-camera annotation · HO-3D multi-camera annotation | mean hand-joint error (cm): 1.17 (±1.12) -> 1.04 (±0.43); 2.22 (±1.22) -> 1.04 (±0.43) | Helped |
| add the 3D distance term (ablation) | HOnnotate multi-camera annotation · HO-3D multi-camera annotation | mean hand-joint error (cm): 1.04 (±0.43) -> 0.98 (±0.40) | Helped |
| add the physical plausibility term (ablation) | HOnnotate multi-camera annotation · HO-3D multi-camera annotation | mean hand-joint error (cm): 0.98 (±0.40) -> 0.99 (±0.40) | No clear effect |
| add temporal consistency (ablation) | HOnnotate multi-camera annotation · HO-3D multi-camera annotation | mean hand-joint error (cm): 0.99 (±0.40) -> 0.92 (±0.34) | Helped |
| add multi-frame joint refinement (ablation) | HOnnotate single-camera annotation · HO-3D single-camera annotation | hand average mesh error (cm): 0.91 -> 0.77 | Helped |
| add multi-frame joint refinement (ablation) | HOnnotate single-camera annotation · HO-3D single-camera annotation | object average mesh error (cm): 0.52 -> 0.45 | Helped |
| add joint direction predictions alongside 2D joint locations (ablation) | CPM + MANO fitting · HO-3D test set | mesh error (cm): 1.14 -> 1.06 | Helped |
| add joint direction predictions alongside 2D joint locations (ablation) | CPM + MANO fitting · HO-3D test set | joint error (cm): 3.14 -> 3.04 | Helped |
| predict 2D joints and fit MANO instead of directly predicting MANO parameters | single-image hand pose estimation · HO-3D test set | mesh error (cm): 1.10 -> 1.06 | Helped |
Closest papers by meaning
- HO-Cap: A Capture System and Dataset for 3D Reconstruction and Pose Tracking of Hand-Object Interaction (NeurIPS 2025 · similarity 0.89)
- HOLD: Category-agnostic 3D Reconstruction of Interacting Hands and Objects from Video (CVPR 2024 · similarity 0.89)
- ArtiBoost: Boosting Articulated 3D Hand-Object Pose Estimation via Online Exploration and Synthesis (CVPR 2022 · similarity 0.89)
- EasyHOI: Unleashing the Power of Large Models for Reconstructing Hand-Object Interactions in the Wild (CVPR 2025 · similarity 0.87)
- How Do I Do That? Synthesizing 3D Hand Motion and Contacts for Everyday Interactions (CVPR 2025 · similarity 0.87)
- SiMHand: Mining Similar Hands for Large-Scale 3D Hand Pose Pre-training (ICLR 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.