New method creates dynamic head avatars with realistic hair

Researchers have developed DynHair, a new method for creating dynamic head avatars with realistic, strand-based hair. This technique uses structured 3D Gaussian Splatting to reconstruct hair from video input, incorporating a temporal network that models strand deformations based on head movement and gravity. The system jointly optimizes the hair's motion and appearance with the face region, resulting in animatable avatars with controllable hair dynamics. Experiments show DynHair achieves state-of-the-art performance in hair dynamics, temporal consistency, and generalization. AI

IMPACT Enhances realism and controllability in virtual avatars, potentially impacting metaverse and gaming applications.

RANK_REASON Academic paper detailing a new method for 3D avatar hair modeling. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• 3D Gaussian Splatting
• DynHair
• Vanessa Sklyarova

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →