Implementation of convolutional sparse coding in deep networks frameworks

529 views Asked by At

I wanted to implement some convolutional sparse coding procedure similar to one described in this paper : http://cs.nyu.edu/~ylan/files/publi/koray-nips-10.pdf I tried with different frameworks (caffe, eblearn torch) but there seems to be lack of tutorials/support for unsupervised feature learning procedures such as this one. The authors say that this particular article is done using eblearn but I found no unsupervised learning procedure there. Have anyone tried to implement these kind of algorithms, and if so which libraries/frameworks did he use? thx

1

There are 1 answers

0
A. Eckert On

I'm trying to do the same. So far i have found a matlab toolbox available at http://www.matthewzeiler.com/software/ (download link at the bottom). 'Convolutional sparse coding' is called 'Deconvolution' by him. The toolbox works, but you have to modify a little bit of code, some matlab functions were renamed.