Great work at this repo.
I was reading the usage on sample.py file and if I am correct one has to provide a .yaml with some of the parameters that make a surface/object (like abc dataset that has num_surfaces or num_edges in the eval_config.yaml) and thus your pipeline will do its magic.
Is it possible to have as an input a common polygonal mesh representation of an object, such as OBJ or STL for such sample.py? or how about if it is directly a point cloud (assuming the cloud is clean enough), with some extra property (like a color) that will help in deciding what points are in certain shape?
Thanks.
Great work at this repo.
I was reading the usage on
sample.pyfile and if I am correct one has to provide a.yamlwith some of the parameters that make a surface/object (like abc dataset that hasnum_surfacesornum_edgesin theeval_config.yaml) and thus your pipeline will do its magic.Is it possible to have as an input a common polygonal mesh representation of an object, such as OBJ or STL for such
sample.py? or how about if it is directly a point cloud (assuming the cloud is clean enough), with some extra property (like a color) that will help in deciding what points are in certain shape?Thanks.