Skip to content

Latest commit

 

History

History
26 lines (14 loc) · 1.11 KB

README.md

File metadata and controls

26 lines (14 loc) · 1.11 KB

Sparse Subspace Clustering (SSC) ADMM Python Implementation with CUDA support

This is a reimplementation of the paper Sparse Subspace Clustering: Algorithm, Theory, and Applications with CUDA support.

Some of the implementation is based on the Matlab SSC-Basis Pursuit ADMM code from the author's website.

The SSC implementation in ssc.ssc.sparse_subspace_clustering.

The corresponding ADMM implementation in ssc/admm.py(CPU) and ssc/admm_gpu.py(CUDA).

The experiment from the original paper is in face_clustering.py and motion_segmenation.py

Install

Please install follow the instructions in CuPY install guide. According to your CUDA version, you will use pip install cupy-cuda12x or pip install cupy-cuda11x.

Other requirements are in requirements.txt.

Dataset

The following datasets should be placed in the ./data folder to reproduce the experiments.

  • Hopkins155
  • YaleBCrop025.mat is in the original Matlab code.