I built a browser-based digital human that listens to a user and responds with generated, lip-synced video. I worked on this project 2 years back, I would like share the learnings with the community.

I’d appreciate feedback from anyone who has worked on real-time avatars, streaming TTS, or interruption handling.

INteresting stuff. What's the next set of featues you intend to add?

Thank you, I plan to make the video generation more realistic