I am currently working on a project where I need to apply SCANnotate to a point cloud dataset that I have already classified using RandLA-Net. My point cloud data is derived from RGB images, and I have all the pose matrices for the camera poses as well as the depth images.
My main issue is understanding how to create the required PKL file to make it compatible with SCANnotate, similar to the preprocessed ScanNet information.
I have reviewed the documentation but am struggling to understand the exact structure and content needed for the PKL file. Could you provide guidance or an example of how to construct this PKL file? Any help or pointers to relevant parts of the documentation would be greatly appreciated.
I am currently working on a project where I need to apply SCANnotate to a point cloud dataset that I have already classified using RandLA-Net. My point cloud data is derived from RGB images, and I have all the pose matrices for the camera poses as well as the depth images.
My main issue is understanding how to create the required PKL file to make it compatible with SCANnotate, similar to the preprocessed ScanNet information.
I have reviewed the documentation but am struggling to understand the exact structure and content needed for the PKL file. Could you provide guidance or an example of how to construct this PKL file? Any help or pointers to relevant parts of the documentation would be greatly appreciated.
Thank you!