Researchers have developed MOCHI, a two-stage framework designed to enhance noisy data from collaborative human-object interaction (MHOI) scenarios. The system first optimizes hand grasps for physical plausibility and semantic consistency with body pose, extending these into full interaction sequences. Subsequently, a diffusion-based noise optimization framework refines the motion of all participants using single-person motion priors, incorporating interaction information. MOCHI has demonstrated effectiveness across various MHOI datasets and applications, including keyframe-based creation and data augmentation. AI
IMPACT Enhances data quality for human-object interaction modeling, potentially improving AI systems that rely on such data.
RANK_REASON The cluster contains a research paper published on arXiv detailing a new framework for enhancing MHOI data.
- alphaXiv
- arXiv
- CatalyzeX Code Finder for Papers
- Connected Papers
- cs.CV
- DagsHub
- Gotit.pub
- Hugging Face
- Litmaps
- Mhoira Leng
- MOCHI
- ScienceCast
- scite Smart Citations
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
