Closed cviviers closed 9 months ago
In this CVPR work, we focused on the pose refinement part, we therefore didn't release the part for initial pose estimation. If you are interested in an end-to-end implementation, you could refer to previous works such as PVNet, PoseCNN, etc.
Recently, we developed a more efficient end-to-end solution, but we still have not released this part as this new paper is still under review. Please stay tuned.
Clear, thanks for the response. Looking forward to the next work!
Hi there. Thanks for sharing your work and all the code. I was wondering if you have an end-to-end implementation available? As in from new input image, initial pose estimation and then the refinement model. From what I can deduce, this repo picks up with a set of precomputed poses already available.
Thanks for any feedback!