Research map / Video / Video segmentation and tracking
Learning to Sit: Synthesizing Human-Chair Interactions via Hierarchical Control
Yu-Wei Chao, Jimei Yang, Weifeng Chen, Jia Deng
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 |
|---|---|---|---|
| Replace a holistic fixed motion with learned reusable subtasks and a freely switching meta controller | Human-chair sitting · Easy | success rate (%): 0.00 -> 31.61 | Improved |
| Use a freely selecting meta controller instead of a predefined walk-turn-sit order | Hierarchical sitting controller · Easy | success rate (%): 25.16 -> 31.61 | Improved |
| Add both left- and right-turn subtasks instead of restricting turning to one direction (ablation) | Hierarchical sitting controller · Easy | success rate (%): 29.38 -> 31.61; 23.01 -> 31.61 | Improved |
| Add curriculum learning from easier to harder starting zones (ablation) | Meta controller · Zone 2 | success rate (%): 0.00 -> 10.01 | Improved |
| Add curriculum learning from easier to harder starting zones (ablation) | Meta controller · Zone 3 | success rate (%): 4.05 -> 7.05 | Improved |
| Fine-tune subtask policies on starting poses sampled from another subtask's ending poses (ablation) | Turn controller initialized from walk · Turn initialized from walk | success rate (%): ~~0.09 -> 51.12 | Improved |
| Fine-tune subtask policies on starting poses sampled from another subtask's ending poses (ablation) | Sit controller initialized from turn · Sit initialized from turn | success rate (%): 32.94 -> 87.41 | Improved |
Closest papers by meaning
- InterMimic: Towards Universal Whole-Body Control for Physics-Based Human-Object Interactions (CVPR 2025 · similarity 0.87)
- NIFTY: Neural Object Interaction Fields for Guided Human Motion Synthesis (CVPR 2024 · similarity 0.86)
- Universal Humanoid Motion Representations for Physics-Based Control (ICLR 2024 · similarity 0.86)
- Hierarchical World Models as Visual Whole-Body Humanoid Controllers (ICLR 2025 · similarity 0.86)
- InterControl: Zero-shot Human Interaction Generation by Controlling Every Joint (NeurIPS 2024 · similarity 0.85)
- DynaMo: In-Domain Dynamics Pretraining for Visuo-Motor Control (NeurIPS 2024 · 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.
