Repository files navigation

⚠️ 🚨 this code base is no longer maintained 😕

A minimalist pytorch implementation of: "Deep Geometric Functional Maps: Robust Feature Learning for Shape Correspondence"[1], appeared inCVPR 2020.

This implementation runs on python >= 3.7, use pip to install dependencies:

pip3 install -r requirements.txt

The preprocessing code will be added later.For the moment, we refer the reader to theoriginal implementation of GeomFmaps to download the data and the preprocessing code.

It should be noted that for each dataset (faust, scape, etc), this module expect that the dataset folder contains 3 folders:

• off folder: this folder contains the meshes
• spectral folder: this folder contains the laplace beltrami related data. It's composed from files having the same name as theoff folder. Each fileis a.mat contaning adict containing three keys:evals,evecs andevecs_trans. This files are created by the preprocessing code.
• corres folder: this folder contains the ".vts" files necessary for the calculation of the ground truth maps.

Use theconfig.yaml file to specify the hyperparameters as well as the dataset to be used.

Use thetrain.py script to train the GeomFmaps model.

python3 train.py

[1]Deep Geometric Functional Maps: Robust Feature Learning for Shape Correspondence