What Object Detection model is better or most used with embedded Nvidia devices (Jetson Xavier NX)?


Hi, I am building a project and I’m using Jetson (specifically a Jetson Xavier NX) as device.

I want to ask to the community a few questions:
· what is the most used Object Detection model architecture (and implementation) for such devices?
· what is the most accurate with detection with a good frame-rate (near real-time) on such devices?

Typically my workflow is: train with Keras/Tensorflow, convert the model with TensorRT and use the optimized engine to make inference on embedded device, so the Object Detection model should be portable with TensorRT.


TensorRT Version: 8
Operating System + Version: Ubuntu 18.04 / Ubuntu 20.04
Python Version (if applicable): >= 3.6
TensorFlow Version (if applicable): 1.15 or 2

This looks like a Jetson issue. Please refer to the below samlples in case useful.

For any further assistance, we recommend you to raise it to the respective platform from the below link