search

robotgradient / grasp_diffusion

Pytorch implementation of diffusion models on Lie Groups for 6D grasp pose generation https://sites.google.com/view/se3dif/home
MIT License
234 stars 24 forks source link

Pre-processing helpers/scripts #2

Closed kuldeepbrd1 closed 1 year ago

kuldeepbrd1 commented 1 year ago

Hi again @TheCamusean

It's super convenient that you provide pre-processed data to train/demo for the Mugs. The training works smoothly with this training data. However, I was hoping to test se3dif on other objects.

Would it be possible for you to upload the pre-processing scripts to generate the pointclouds from object meshes and final grasp annotations? Or point out how I could go about it after obtaining water-tight meshes using https://github.com/hjwdzh/Manifold ?

Thanks a lot in advance

robotgradient commented 1 year ago

Thanks for you inquiry @kuldeepbrd1 .

We are working on a more complete repository with additional features as data generation pipeline and we expect to upload them for the end of the week.

I keep you updated.

kuldeepbrd1 commented 1 year ago

That's Awesome :)
Looking forward to it.

robotgradient commented 1 year ago

@kuldeepbrd1 You might find the dataset generation in the dev branch in the scripts folder.

We will move it to the master branch soon.

kuldeepbrd1 commented 1 year ago

I see the dev branch. That is already great for me to start playing with it.

Cheers!