Of course you could, but that would be for a different project. The goal here was to have it in the terminal. I you just want it try this: https://jeeliz.com/demos/faceFilter/demos/threejs/matrix/

oh it's even on the todo. Let me add that in a bit

Shoul be available now pip install matrix-webcam --upgrade and the matrix-webcam -d 0[1,2,3,...]

Awesome! It works now! Thanks. Device is selected by an actual number, not the device path, but easy enough to figure out. I guess the help actually says that, if I paid attention...

exactly, see the README on how to do it. I am not taking responsibility if you do it with HR though

Could anyone provide better instructions? I'm new to OBS and have spent almost an hour trying variations of things with no success. The matrix-webcam runs in my terminal, but it outputs whatever OBS sees (video feedback infinite fractal). README.md instructions leave too many steps out.

I assume the method is run in a window > capture window in OBS > OBS virtual camera to Zoom / whatever.

It uses mediapipes selfie segmentation module: https://google.github.io/mediapipe/solutions/selfie_segmenta... using MobileNet v3 (around 0.5 GFlops of juice it's gurgling, so enough to run on a raspberry, maybe not your Casio)

It'd be interesting if this could be embedded in the client/server, e.g. Jitsi. At work we turn off cameras to save bandwidth, but this seems like you could just send a bitstring, and heavily compressed at that.

Actually that reminds me of - and i personally think it's the future of video compression - that there a neural network based codecs in the making https://developer.nvidia.com/ai-video-compression

Yeah would be fun if you could gzip/compress let's say 4 frames every 200ms and make your own ascii video stream codec so to speak