Running Object Detection with object Tracking on Jetson Nano

AI & Data Science Deep Learning (Training & Inference) TensorRT
1

Jetson Nano Specification:
Hardware Platform (Jetson / GPU): Jetson Nano 4 GB RAM
JetPack Version : 4.6.1
TensorRT Version: 8.2.1.8
NVIDIA GPU Driver Version (valid for GPU only):10.2
Issue Type( questions, new requirements, bugs): question

Hii, Team Nvidia:

I am trying to run multi object detection and tracking on jetson nano in real time.
As per current approach I am using Yolov5s (optimized with tensorrt) for object detection and SORT for object tracking. However, I am not getting a desired result. The average FPS is aroung 6-7 given 4-5 objects in frame. And the tracker does not work well.

I need suggestions:

  1. How can I get a real - time fps (average 20) with the current approach?
  2. Is there any other tracking algorithm we can utilize to do object tracking?

Thank You.

2

I am moving this topic to the TensorRT forum for better visibility.