Blogs like this make me wish I would have gone for a PhD, perhaps in imagery exploitation. This one comes from a friend of mine from the Air Force Research Laboratory now teaching at the U.S. Air Force Academy. Anyone who knows what I do for a living would understand why I find this blog interesting. Read the original blog from my buddy, D.O.G., if you love math.
It’s been far too long since my last blog post, but now that the linear algebra class I’m teaching is coming to an end, I hope to update a bit more.
This one is a long time coming, considering I blogged about the main idea almost 6 months ago. But that blog entry only described our approach in the noiseless case, whereas the bulk of the paper actually tackles stability in the noisy case. This is an important stride for phase retrieval, considering the only provably stable alternative is PhaseLift/PhaseCut, while our approach is completely different.
To review, here’s the main idea: We have a collection of vectors that we use to measure our signal. If we could determine the phases of these measurements (up to a…
View original post 887 more words