✨SyncMV4D: Synchronized Multi-view Joint Diffusion of Appearance and Motion for Hand-Object Interaction Synthesis

📝 Summary:
SyncMV4D generates realistic and consistent multi-view 3D Hand-Object Interaction videos and 4D motions. It unifies visual priors, motion dynamics, and multi-view geometry, using a joint diffusion model and a point aligner for robust generation.

🔹 Publication Date: Published on Nov 24

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2511.19319
• PDF: https://arxiv.org/pdf/2511.19319
• Project Page: https://droliven.github.io/SyncMV4D/
• Github: https://droliven.github.io/SyncMV4D/

==================================

For more data science resources:
✓ https://t.me/DataScienceT

#HandObjectInteraction #DiffusionModels #3DGeneration #ComputerVision #GenerativeAI