TensorRT-For-YOLO-Series

Robotics & Edge Computing Jetson & Embedded Systems Jetson Orin Nano
1

Hello!

I created an onnx file from a pth file and a trt file according to the README of “TensorRT-For-YOLO-Series”.

I executed the command written in the README:
python trt.py -e yolox_s.trt -i src/1.jpg -o yolox-1.jpg --end2end
1.jpg is an image with a person in it.
yolox-1.jpg is the image with the inference result.
However, yolox-1.jpg is the same as 1.jpg, and it seems that nothing was detected.
Where could the problem be?

3

Hi,

Do you use the default testing image or a custom one?
Could you try to use the image attached in GitHub to see if the accuracy is good?

Thanks.

4

Hello.
I tested with the image available on GitHub. I used 1.jpg, but the output image was the same as 1.jpg.

input：

1
11280×720 113 KB

output：
yolox-1
yolox-11280×720 162 KB

Is there a problem with the trt file I created following the README on the GitHub of TensorRT-For-YOLO-Series?

Thank you in advance.

5

Hi,

Do you mean any error when converting the engine into TensorRT?
If not, please check with the library owner to see if any extra processes are required to run the app on Jetson.

Thanks.

7

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