The following is the backend code for the lipreading project, maintained at this repo.

Features of the project:

• Receive live videoframes via websockets

• Annotate the frames to provide "almost" real-time speeds of lip reading.

Dependencies are in the requirements file, it is suggested to couple this with ngrok for remote access as well.