Optimum way to design custom parallel deepstream flow

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU)
Nvidia GPU
• DeepStream Version
• JetPack Version (valid for Jetson only)
Jetson NX
• TensorRT Version
what ever in DS6.3
• NVIDIA GPU Driver Version (valid for GPU only)
Driver Version: 535.171.04 CUDA Version: 12.2
• Issue Type( questions, new requirements, bugs)
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)


Just wondering what is the best/optimum way of doing this C++ deepstream application. I have got the first and second dataflow working without any issue, I am wondering if the third dataflow consists of sgie to a message broker is an optimum solution. The sgie I am using probe to skip, filter on particular stream, filter on objects before sending a custom message to my python app. I am communicating to my python app using zmq via the msgbroker (Pub/Sub). The dsvideowriter is a custom module based on dsexample. The whole application works with the first two dataflows talking to my python app. I just need to add this custom sgie and I can’t think of a optimum way of doing this.


decided to create a custom nvinfer to simplify pipeline

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