Datasets: All experiments were conducted on standard benchmark datasets, including FAUST, SCAPE, SMAL, DT4D, and TOPKIDS, obtained from their official sources.

Environment Setup: For the required dependencies, please refer to the implementation of "Hybrid Functional Maps for Crease-Aware Non-Isometric Shape Matching" (which utilizes the elastic basis).

Project dependencies:

cuda: 11.7 python: 3.8.19 pytorch: 2.2.2 h5py 3.11.0 numpy 1.22.3 pip 23.3.1 torch_cluster 1.6.3+pt22cu121 torch-geometric 2.6.1 torch_scatter 2.1.2+pt22cu121 torch_sparse 0.6.18+pt22cu121 trimesh 4.3.1