search

NVlabs / Deep_Object_Pose

Deep Object Pose Estimation (DOPE) – ROS inference (CoRL 2018)
Other
1.03k stars 287 forks source link

need help for coding #240

Closed kooshyarkosari closed 2 years ago

kooshyarkosari commented 2 years ago

Hello

I have downloaded the Ycb-video Dataset, I want to read the images directly from the file on my PC then run the DOPE and see the poses of the object instead of using cameras, ROS, and so on, which part of the code must be modified? ( detector, inferences,... ) How can I change it? There is any sample code for this task?

Thanks in advance

TontonTremblay commented 2 years ago

https://github.com/NVlabs/Deep_Object_Pose/blob/master/scripts/train2/inference.py check this file. This is inference without ros.

TontonTremblay commented 2 years ago

After that I think in one other issue I shared how I computed the ADD. But you can use the code Yu wrote.