Media Streams for AI Processing

I wanted to create a AI based proctoring application on Openvidu and my aim is to send the media stream from media server to the receivers and also a Python ML service which raises alerts on malpractice

Hello @shah,

Currently the only way to export the stream from an OpenVidu session is recording it and then processing it offline.

Maybe in the future we can evaluate to export the media with RTP to be processed with ffmpeg or similar tool.

Best regards

1 Like

Hello @micael.gallego ,

Thanks for the reply.
Is there any way to send by media stream to another endpoint for processing in real time using kurento?
It need not necessarily be an openvidu setting. Can I maybe edit something in kurento while using openvidu to achieve the same?

Thanks and Regards