preprint · arXiv · Jan 2, 2026
Avatar Forcing: Real-Time Interactive Head Avatar Generation for Natural Conversation
Taekyung Ki, Sangwon Jang, Jaehyeong Jo, Jaehong Yoon, Sung Ju Hwang
In plain English
Most talking-head systems only speak; this one reacts. The authors propose a diffusion framework with preference optimization that lets an avatar respond to a user's speech and gestures with minimal delay, without extra labelled training data.
talking head generation
real-time avatars
diffusion