I want to use PeopleNet for detecting People on a Camera Stream. The Problem is that i have to integrate it into an existing Python Application and I only get the current frame as an opencv-image. Thats why I can’t use the “full” Deepstream Pipeline, i only need the part for inferencing.
What Code do I need to get the PeopleNet Predictions for a camera frame in an existing application?
Thank you so much for your help, I love the Jetson Devices and the prebuilt models you guys are providing.