Sparsenet-conv is a project mainly written in OBJECTIVE-C and MATLAB, it's free.
Convolutional sparsenet! Author: Jack Culpepper License: BSD
This program is related to the following papers:
B. A. Olshausen and D. J. Field. Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature, 381(6583):607-9, jun 1996.
B. A. Olshausen BA (2003). IEEE International Conference on Image Processing. Sept. 14-17, 2003. Barcelona, Spain.
Start matlab; startup.m will run automatically and set up your path.
To test, run unittest.m .
To check gradients, run:
check_a_conv.m check_phi.m
To collect all the individually saved basis function images into a movie, edit fig_avi.m and set paramstr to the paramstr you would like to watch evolve, then run fig_avi.m .
--bjc 2010-03-10