We study convex relaxation algorithms for phase retrieval on imaging problems. We show that structural assumptions on the signal and the observations, such as sparsity, smoothness or positivity, can be exploited to both speed-up convergence and improve recovery performance. We detail experimental results in molecular imaging problems simulated from PDB data.Let us note that this phase retrieval approach is part of the inverse section of the Scattering Transforms and Applications with an attendant explanation as to why:
The Implementation of PhaseCut in Octave/MATLAB is here. I also note the following important element when it comes to reproducible research: the ability to run it on bare bones matlab or octave:
Our toolbox works on all recent versions of MATLAB on Mac OS X, and on MATLAB versions anterior to 2008 on Linux.
A big thank you for Fajwel, Irène and Alexandre for implementing their toolbox and minding about this issue.
Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.