search

THU-DA-6D-Pose-Group / GDR-Net

GDR-Net: Geometry-Guided Direct Regression Network for Monocular 6D Object Pose Estimation. (CVPR 2021)
https://github.com/THU-DA-6D-Pose-Group/GDR-Net
Apache License 2.0
277 stars 47 forks source link

inference real-time during capture rgb #118

Open stupidyoh opened 10 months ago

stupidyoh commented 10 months ago

Hello!

After the last question, I succeeded in learning my own dataset. And then I'm trying to inference 6d pose of target object during capturing rgb image. I'm referring to "plot_prediction_multi.py" and "do_test" function of "engine.py" code. But it seemed to infer the result only for the saved and annotated image.

My question is "Is it possible to estimate the 6D pose of the target object just in the new rgb image?" and "Is annotated information necessary to estimate the pose of the target object in the new test image?".

My goal is to get predicted 6d object pose(or 3d bounding box) captured by camera for about 30~60 fps. If so, can I ask for some advice on which codes to fix?

Thank you for your help!

xwong22 commented 9 months ago

Hi @wangg12 , is there any updates on this issue? I would like to use the pretrained model for ycbv dataset for real-time inference on a ycb object