Following up on Imaging With Nature: Compressive Imaging Using a Multiply Scattering Medium (but also [2,3] ) and after I probably spammed him, Jared Tanner sent me the following:

Dear Igor,

Thank you for alerting me to your work. I’m so glad that you have turned your vast expertise from the blog into research results. And in particular congratulations on its being published in nature. In reading it I noticed that you are using an MMV framework from three illuminations. Though MMV isn’t a topic I am an expert on, I have stumbled upon an interesting feature that the CGIHT algorithm benefits more from MMV than do other competing algorithms, see Figure 5 of

http://people.maths.ox.ac.uk/tanner/papers/BlTaWe_cgiht.pdf

Note that we haven’t compared with OMP so don’t know how well it might be performing.

All the best,

Jared

-------

Jared Tanner

Professor of the Mathematics of Information

Mathematics Institute

University of Oxford

http://people.maths.ox.ac.uk/tanner

To which I replied:

Also, with regards to MMV, Jeff [Blanchard] had alerted me to some of these results back in January and I even blogged about it.

http://nuit-blanche.blogspot.fr/2014/01/sharp-phase-transition-for-joint-sparse.html

I have to say that I should probably have pushed for y'all preprint to be used as a reference but the paper had been written several times already and I was getting fed up with the different submission processes (I will tell that story eventually).Please note that MMV has provided interesting results in a similar field: http://nuit-blanche.blogspot.fr/2013/10/imaging-strong-localized-scatterers.html

**Join the CompressiveSensing subreddit or the Google+ Community and post there !**

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.

## No comments:

Post a Comment