I am a researcher in dialog system field. We built a system using Freeswitch as the front to collect conversation videos.
I have a question about the Freeswitch video recording component.
Freeswitch streams the video and saves it incrementally. I need to extract the head position of the human in the video in real time. I have a head tracker that can access the video stream from a webcam and process the video stream from the webcam in real time. But I am not sure how to stream the incrementally recorded video from Freeswitch to my head tracker. Could Freeswitch broadcast the video to somewhere else? Or Could you point me to where I can modify Freeswitch to integrate my head tracker in?
Another small question is can the audio be broadcast the same way?
Thanks a lot!