3

So I have to use MS Teams for work. Thankfully there is a linux client, but it's missing a few features. I'm looking simulate the background blur effect, but I'm not quit sure where to begin.

Teams only sees the device in the laptop. I'd like to capture webcam device /dev/video0 process it to include the background effect and then present a second device that MS Teams could use.

Seems like ffmpeg and gstreamer can provide output streams and the inline hooks to process the video, but Teams wants a device. Bonus points for including audio effects ;)

Any ideas?

Improve this question
3
  • See Is there any way ffmpeg send video to /dev/video*?
    – llogan
    Commented Apr 23, 2020 at 2:27
  • okay, v4l2loopback connects the dots between the existing webcam and the new virtual device that would be used by MS Teams. Now I'm looking into the facial identification. The ffmpeg filters have lots of static filter options, but all the content I'm finding is about some form of static blurring. I know half the problem of finding something is knowing what to ask for. Are there any projects out there or filters that can track a subject in a video feed and dynamically apply a mask or send coordinates to the blur filter?
    – Pedulla
    Commented Apr 23, 2020 at 16:25
  • Not with ffmpeg alone. Maybe OpenCV? Or Cheese. Not an area I have any experience with.
    – llogan
    Commented Apr 23, 2020 at 17:19

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .