Order of GIEs in deepstream pipeline

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) Jetson Orin AGX 32GB
• DeepStream Version 6.2
• JetPack Version (valid for Jetson only) 5.0.1
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, bugs)

I have a question about the order of the GIEs in deepstream pipeline.
My application is trying to use pgie(detector) - sgie 1(detector) - sgie 2(classifier).
The sgie1 will infer the output from pgie, and the sgie 2 will infer the output of the sgie 1.
Like the deeptream-app-test2, I am thinking of pgie - tracker - sgie - sgie but since the example’s sgies infer the output of the pgie, I need some advice.
Is is recommended to put the tracker after the pgie and the sgies follows? Like the below pipeline.
PGIE - Tracker - SGIE - SGIE

3

The pipeline is based on your use case. But this is the typical use case pipeline as some SGIE depend on the object id which generated in tracker: PGIE - Tracker - SGIE - SGIE

1 Like
4

Thanks :)

5

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.