Deepstream sample for running with primary and secondary inferences

I am developing an application with two inferences, primary and secondary.

Bounding boxes (labels: box1 and box2) are extracted from primary inference and fed into secondary inference.
But I need to do a bit of processing at bounding boxes before fed into secondary inference.

To make sure box2s are inside box1s and only box2s are fed into secondary inference.

What application should I follow?

Currently I am using deepstream-test3, I can extract bounding boxes after primary inference. I can check box2s are inside box1s and get ready to feed into secondary inference.

(1)But I don’t know how to add in secondary inference into the pipeline.

I like to run maximum speed. So I am worried that box selection blocks the pipeline. (2)What could be the best way to handle this?

You can check test2 to see hwo to add secondary inference.
You can add a gstreamer plugin as dsexample between pgie and sgie if it’s a compute-intensive work.

1 Like

Thanks let me check.