search

paulpanwang / POPE

Welcome to the project repository for POPE (Promptable Pose Estimation), a state-of-the-art technique for 6-DoF pose estimation of any object in any scene using a single reference.
https://paulpanwang.github.io/POPE/
134 stars 11 forks source link

Estimate the 6DoF Object Pose #3

Open dingdingcai opened 1 year ago

dingdingcai commented 1 year ago

Hi,

Thanks for the nice work!

I noticed that only the relative 3D rotation accuracy is evaluated and reported in the paper, how about the relative 3D translation? Is it possible to estimate the full 6DoF pose using POPE?

YangHai-1218 commented 1 year ago

Hey, I have the same question.

paulpanwang commented 9 months ago

Indeed, you have a keen perception. In the scenario where we only have a single reference image, we can only estimate the scale based on the object's bounding box size in the 2D image (following the approaches of Gen6D and RelPose++). Training a scale regression network on a dataset that encompasses real-world scale relationships may be a viable solution.