Research Seminars

Sparsifying sub-band decompositions
Dr. Mike Davies, DSP Lab, Queen Mary University of London
Wednesday 4 Feburary 2004

We present a solution for constructing overcomplete sparse subband decompositions. This is a generalization of the perturbed basis pursuit problem specifically applied to an overcomplete subband representation. Our formulation is based upon the iterative re-weighted least squares algorithm and can be given a probabilistic interpretation. Although the convergence properties of this algorithm are known to be slow, we observe experimentally that only a few iterates are sufficient to generate a reasonably sparse approximation. Furthermore using subband bases provides us with an algorithm whose complexity grows linearly in time.