Hey, I want to receive each individual user recording stream to apply face detection. Is it possible through jibri?

I want to receive the each connected participant stream through jibri and want to apply Machine Learning on each client stream.

The very purpose of Jibri is to do the opposite, that is, generate a composite media from the individual video/audio streams. It launches a browser in a X11 server, does a x11grab of the screen and feeds it to ffmpeg.

That’s means it gives a combined stream not like array or collection of streams for individual user.

Check the followings