How to use DeepSteam to separate video and audio streams?

Xavier and DS 4.0

I got a rtsp stream which contains video and audio. My problem is how to separate them, and do some analytics further (eg. face recognition + speech recognition )?

DeepStream SDK is to demonstrate the advantage of using Jetson platforms on Deep Learning Inference. You can run deepstream-app to get hardware acceleration in video decoding, inferencing, frame-resizing and on screen display. For audio part, you would need to look for 3rdparty solution and integrate into deepstream-app.