search

vimalabs / VIMA

Official Algorithm Implementation of ICML'23 Paper "VIMA: General Robot Manipulation with Multimodal Prompts"
MIT License
778 stars 87 forks source link

Settings for Mask RCNN version #1

Closed kenoharada closed 2 years ago

kenoharada commented 2 years ago

Thank you for sharing your work! I really enjoy playing with VIMA.

Your code in example.py assumes ground truth segmentation mask of object is given.

Can you share codes for Mask RCNN version where segmentation mask is not given by simulator? Or can you tell me which kind of modification is needed for Mask RCNN version?

yunfanjiang commented 2 years ago

Hey, thanks for your interest! We are currently working on clean-up and will release corresponding code soon.

kenoharada commented 2 years ago

Thank you!

kenoharada commented 2 years ago

@yunfanjiang training code is going to be public in the future?

yunfanjiang commented 2 years ago

Because the training code is tightly coupled with company infra, I'm afraid that we will not be able to disclose it. Apologize for that. But we will let you know if anything changes.

kenoharada commented 2 years ago

OK, thanks!