Hi,I saw a new feature added in react module of jitsi meet files in github which is facial-recognition,I had built the jitsi-meet android app but the facial recognition feature wasnt working,what might be the issue or how to activate it?
Hello! At this moment the facial-recognition/facial-expressions feature is only working on web. You can test it on your local instance by setting the enableFacialRecognition flag to true. You can do that by adding
#config.enableFacialRecognition=true to the url.
Where I can find more details and use cases?
Yes, what are the use cases? How does this work? What are the security implications? The doc note below is very cryptic. Please explain:
I’m glad this is whitelisted. I’m setting to false right now until a good justification of it’s use can be explained.
the screenshot here pretty much says it all. It’s a tool for speakers to evaluate the apparent effect of their speech on listeners. The very idea of an organisation where this kind of feature could be considered useful makes me laugh (unfortunately the tool seems to have difficulty to recognize it actually)
This begs the question: How would participants feel if they knew their video was being fed through facial recognition software no matter how innocuous it may be.
Imagine video conference 120 years ago, in Switzerland.
At the end of a Physik course, the teacher says sternly: ‘Young Einstein ! Der Rechner says that you smiled no less than 174 times during my course on the ether ! Can you explain what you did find so funny ? No ? Is it too complicated for us maybe ? I’m afraid you never will come to anything’.
Crushed. Poor Albert would feel crushed.
Does this feature work on encrypted sessions?
I think you are missing my point. With respect to video conferencing software, “Facial Recognition” can be a red flag. It’s all about media optics.
This feature is completely client-side, no server interaction is required, so it doesn’t matter what encryption is enabled.
No worries then
Ohh,Thank you very much,I would love to see the feature for mobile app too.
Can I get an idea about how and from where you received the video stream for passing it to the face-api.js for facial recognition,I would like to build a survillance system using jitsi so it would be helpful for my project.