Description
In this page, we include the Matlab code associated with the paper G. Silva, J. Quesada and P. Rodriguez, "Efficient Separable Filter Estimation Using Rank-1 Convolutional Dictionary Learning", IEEE International Conference on Image Processing (ICIP), 2018.
MATLAB code
The matlab code provides a usage example in the demo folder. The core file in the package is sep_F_APG_cbpdndl.m.
[D, Dv, Dh, X, optinf] = sep_F_APG_cbpdndl(D0, S, lambda, opt)
sep_F_APG_cbpdndl_gpu : D0 is the initial dictionary, S represents the training set, lambda is the sparsity parameter and opt is an array that contains fixed values of configuration.