YOLO v4 tiny on Jetson AGX xavier

Autonomous Machines Jetson & Embedded Systems Jetson AGX Xavier
1

Hii all,
I am a beginner working with Jetson nano and Jetson xavier.
Recently I have implemented yolo v4 tiny on jetson nano with webcam using darknet. The model worked well and the webcam fps was nearly 4 fps.
So I have tried it using Jetson xavier so that I can increase the fps.
But surprisingly even on Jetson xavier I am getting the same 4 fps with webcam. Can anyone tell me the reason behind this and how to avoid it??

thanks in advance

3

Hi,

Have you maximized the device performance first?

$ sudo nvpmodel -m 0
$ sudo jetson_clocks

More, it’s recommended to try our Deepstream sample to get the optimal performance:

Thanks.

5

Thanks for replying me

but even after changing the power mode and after using the second command iam still getting the same fps.
is there anything I need to change??
thanks in advance

7

Hi,

Could you try to deploy YOLOv4 Tiny model with TensorRT?
A sample can be found below:

We recommend using TensorRT since the inference implementation has been optimized for Jetson.
Thanks.

9

Thanks for the advice I have used Tensorrt. The FPS is pretty good now
Thanks again

10

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