search

DC1991 / G2L_Net

MIT License
118 stars 13 forks source link

About the training data #8

Closed sunshantong closed 3 years ago

sunshantong commented 3 years ago

Hi, thank you for your great work. Will you provide the completed training dataset? Or how do I generate it myself? Thank you very much.

DC1991 commented 3 years ago

@sunshantong Hi, thanks for your interest in our paper and code. I have described how to generate point-wise label with the object 3D model in the paper (in the experiment section). I borrow the render code from this git (https://github.com/thodan/bop_toolkit): render the 3D object model to the scene with GT pose, and label the point (transformed from depth with known camera parameters) less than distance threshold to the rendered 3D model as positive point.

sunshantong commented 3 years ago

@DC1991 Thank you for your quick reply. Sorry for my negligence. I'll render it in this way. Thank you very much.