Will the code and support for comparative experiment with MPM be open source?

robo-alex / gs-dynamics

Code for "Dynamic 3D Gaussian Tracking for Graph-Based Neural Dynamics Modeling" (CoRL 2024)

https://gs-dynamics.github.io/

36 stars 5 forks source link

Will the code and support for comparative experiment with MPM be open source? #4

Open AlanYoung123 opened 1 week ago

AlanYoung123 commented 1 week ago

Thanks for your open source great work! In the part of action-conditioned video prediction, you compared your method with MPM. You added support for two types of robot end-effectors and optimized the physical property using CMA-ES. Will this part of code be open source? Looking forward to your reply!

robo-alex commented 1 week ago

Thank you for your interest Alan! As we have a concurrent project using this part of the code, we are not going to release it recently. We would like to discuss the details and provide support if you are implementing it. Stay tuned!

AlanYoung123 commented 1 day ago

Thank you so much for your reply Alex! I want to implement this part of code, but I had some problems. I`ll appreciate it if you can give some clues about how to add support for the robot end-effector and determine the rotation angles as well as simulation bounding box as in PhysGaussian and how to apply the "action.txt" file in mpm. Thanks again!