Get data after analysis for Yolo-deepstream

Hi,

I have two questions :

  1. I want to exploit the data that I can obtain after using deepstream-YOLO, data like coordinates of the bounding boxes, the name of the detected object, the threshold, etc.
    Is anyone know where I can find this, where I can understand where to find this?

  2. I will exploit these data for a robot, my goal is to use a jetson nano with deepstream, use the data of the bounding boxes to create a program that will control motors of the robot.
    Do you think deepstream is an appropriate program to help me doing that? Or do I have to change to another program, or add another program that can help me?

I’m a beginner in computer vision and robotics.

Sincerely,

Hi, I ask again, is there anyone who can help me understand where to get the data of the bounding boxes in deepstream?

Hi,
Sorry for late!

You can refer to

  1. https://docs.nvidia.com/metropolis/deepstream/Custom_YOLO_Model_in_the_DeepStream_YOLO_App.pdf
  2. TensorRT/plugin/nmsPlugin at master · NVIDIA/TensorRT · GitHub (because the last layer is NMS)

BTW, we have some other straightforward samples - GitHub - NVIDIA-AI-IOT/deepstream_tao_apps: Sample apps to demonstrate how to deploy models trained with TAO on DeepStream