Auto-AVSR: Audio-Visual Speech Recognition with Automatic Labels

Not referenced in the README, here's a great video demonstration of this type of AVSR network running in real time:

https://m.youtube.com/watch?v=XDO8OYnmkNY&t=120s