About Pose Estimation

[WY-2022]

Hello.

I didn't find the pose estimation part in the code you posted. How does the pose estimation and refinement work after extracting the geometric primitives?

Also, I previewed the dataset you provided online and I found no 6D pose groundtruth in it. How to get them?

Is the pose estimation part learnable?

[WY-2022]

I have read this paper, but the pose estimation part is difficult for me to understand because of my limited knowledge. So I want to learn this part directly by reading the code.

[vrk7]

Where you able to replicate the segmentation part of this paper? @WY-2022

[WY-2022]

@vrk7 I believe that in the section "Implementation of Proposed Architecture" of ART-Net/SourceCodes/Train_of_ART-Net_Stage-2.ipynb, the author provides several code implementations similar to the UNet structure.