Dear all,
I am using the standard Yolo Darknet classes, on the Jeston TX1, and it works fine.
My future goal is to use the same file, just using not the entire classes, but just some on my Jetson TX1.
For example, I just want to train the classes “horse” and “vehicle”, not the rest.
Could you please help me to find the good process to achieve this point ?
Thank you for your help and support.
Regards,
Hi,
Unfortunately, training a networks may not work on the Jetson TX1.
Please noticed that Jetson platform is designed for fast inference as an edge device.
It doesn’t have a large storage and data bandwidth which required by the training.
If you want to give it a try, you can follow this page to train a customized YOLO networks:
But it’s still recommended to apply the training work on a desktop environment.
Thanks.