Research map / 3D vision / Human pose estimation
H2O: Two Hands Manipulating Objects for First Person Interaction Recognition
Taein Kwon, Bugra Tekin, Jan Stühmer, Federica Bogo, Marc Pollefeys
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 right-hand pose input to left-hand pose (ablation) | TA-GCN · H2O interaction recognition (test) | accuracy (%): 33.61 -> 58.92 | Improved |
| add object pose to both hands pose (ablation) | TA-GCN · H2O interaction recognition (test) | accuracy (%): 58.92 -> 79.25 | Improved |
| add hands-object interconnections to no interconnection (ablation) | TA-GCN · H2O interaction recognition (test) | accuracy (%): 75.52 -> 78.84 | Improved |
| add left hand-right hand interconnections (ablation) | TA-GCN · H2O interaction recognition (test) | accuracy (%): 75.52 -> 76.76 | Improved |
| add all interconnections over hands-object only (ablation) | TA-GCN · H2O interaction recognition (test) | accuracy (%): 78.84 -> 79.25 | Improved |
| add Sj term (ablation) | TA-GCN · H2O interaction recognition (test) | accuracy (%): 73.44 -> 79.25 | Improved |
| add Tj,inter term (ablation) | TA-GCN · H2O interaction recognition (test) | accuracy (%): 75.52 -> 79.25 | Improved |
| add Tj,intra term (ablation) | TA-GCN · H2O interaction recognition (test) | accuracy (%): 76.76 -> 79.25 | Improved |
| add Aj,inter term (ablation) | TA-GCN · H2O interaction recognition (test) | accuracy (%): 76.35 -> 79.25 | Improved |
| add Aj,intra term (ablation) | TA-GCN · H2O interaction recognition (test) | accuracy (%): 77.59 -> 79.25 | Improved |
| use pose-based TA-GCN instead of image-based SlowFast | action recognition · H2O interaction recognition (test) | accuracy (%): 77.69 -> 79.25 | Improved |
| use TA-GCN instead of ST-GCN | pose-based recognition · H2O interaction recognition (test) | accuracy (%): 73.86 -> 79.25 | Improved |
Closest papers by meaning
- H+O: Unified Egocentric Recognition of 3D Hand-Object Poses and Interactions (CVPR 2019 · similarity 0.92)
- Transformer-Based Unified Recognition of Two Hands Manipulating Objects (CVPR 2023 · similarity 0.91)
- Collaborative Learning for 3D Hand-Object Reconstruction and Compositional Action Recognition from Egocentric RGB Videos Using Superquadrics (AAAI 2025 · similarity 0.87)
- EgoPressure: A Dataset for Hand Pressure and Pose Estimation in Egocentric Vision (CVPR 2025 · similarity 0.87)
- HOT3D: Hand and Object Tracking in 3D from Egocentric Multi-View Videos (CVPR 2025 · similarity 0.87)
- HO-Cap: A Capture System and Dataset for 3D Reconstruction and Pose Tracking of Hand-Object Interaction (NeurIPS 2025 · similarity 0.86)
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.
