How to prepare Stanford3dDataset for training?

[TomHeaven]

I try to replicate your experiment but fail with the missing data folder error asking for Stanford3dDataset_v1.2_Aligned_Version/graph_v.

I try the data pre-processing steps as recommended in https://github.com/loicland/superpoint_graph, and get three folders: clouds, features, superpoint_graphs. However, what is the connection between the three folders and 'graph_v, graph_p`?

Looking forward to your help.