When I ran object_detection on kaya, I found that I could detect only three kinds of objects. I wondered if I could detect more. I saw the example of yolo in samples.I wonder if I can train my own data to identify more junk, and if so, what are the detailed steps?
Could you please check this part of documentation and let us know if it helps?
http://isaac.nvidia.com/isaac-2019.2-dev/isaac/packages/yolo/doc/yolo.html?highlight=train#training-fine-tuning-the-network
I can’t log on the website you gave me. May I ask whether it is the problem of this website or something?