End-to-end model that listens, sees, thinks and responds on video in real time
twitter.com
1 thread
How is the first token latency for real-time scene processing being addressed?
How is the first token latency for real-time scene processing being addressed?