DARPA Projects

Knowledge enhanced exapixel imaging [KEEP]

The US Defense Advanced Research Agency (KECoM-FP-013)

Dates: 2011-2012 [PhD student: Anastasios Kyrillidis]

In KEEP, we develop a mathematical infrastructure, incorporating explicit structured priors and learned signal and task dependent priors for highly compressive signal measurements and decompressive inference. Our final objective is to demonstrate the capacity of this mathematical infrastructure for greater than 100× compression ratio imaging in wide-field, high-pixel-count imaging.

Team: David Brady-PI, Larry Carin, Mauro Maggioni, Robert Calderbank, Ingrid Daubechies, Volkan Cevher, Stan Osher, Guillermo Sapiro, and WotaoYin.