Dose Deepstream support adding new video source in pipeline?

Suppose we process three videos in the meantime. And now, a new video source need to be processed. To use GPU source as much as possible, we want to add the new video to the process pipeline. Is it possible in Deepstream structure?


You can use the different deepstream pipeline for each input source.

Maybe I don’t state my question clearly. What I want to do is adding the new video into the current pipeline which processes the previous three videos. As a result, the Deepstream could decode four videos in the meanwhile. And it will be much better for the inference since the time of inference will be shorter than processing them separately.

Is it possible for adding new videos or RTSP streams into the existing pipeline dynamically?


If you want to add a source dynamically, please design your mechanism as a “Custom Gstreamer Plugin”.

RTSP source is supported, please check our document for details:

[b]> Reference Application Configuration

Configuration Groups
Source Group[/b]
The URI can be a file, an HTTP URI, or an RTSP live source.


Thank you for your reply. I will have a look at it.