Closed hideonInternet closed 3 years ago
Hi, I only have my research code to offer (which isn't very clean and well factored): shape_pca_generic.py.zip. This assumes that all the shapes to which a PCA model has to be fit, are already saved to disk. But it shouldn't be too hard to interface this with benchmark.py.
Nice work! Could you please provide the PCA code used for analyzing the learned space of the train set? Thanks a lot!