MediaPipe Meet vs. MediaPipe Selfie Segmentation Model (Virtual Background)

Hi,

this topic is about the utilization of MediaPipe Meet and the unclear situation regarding its licence.

As far as i know, the MediaPipe Meet Model is licensed under Google Terms of Use and not Apache 2.0 anymore. So it would make sense to me to replace the current Model with the MediaPipe Selfie Segmentation Model to avoid possible legal issues with Google.

Are there any plans to switch to an other Model in the future?

References:
https://github.com/tensorflow/tfjs/issues/4177#issuecomment-858137878

https://google.github.io/mediapipe/solutions/selfie_segmentation

Best regards

Hi there!

We have the same issue here, don’t wanna get sued by mistake. :wink: Think @Jitsiisgreatbut hat a nice idea, to have it rock solid.

Kind regards

Jitsi-meet use a version downloaded before the license change. So there is no immediate problem to solve.