Version: February 21st, 2016
The aim of the program is to test the efficiency of the Sparse Subspace Clustering technic available at http://www.vision.jhu.edu/ssc.htm.
The code has been adapted to one example of the Hopkins-155 dataset, namely ‘1R2RC_truth.mat’.
The main program follows the following steps:
- Load the data and build the data matrix
- Plot the first frame.
- Plot the first frame with the result of a kmeans.
- Compute the Agency matrix and plot the first frame with the result of a Spectral Clustering.
- Compute the Agency matrix from the Sparse Subspace technic and plot the first frame with the result of a Spectral Clustering.