The Wireless and Video Communications Lab is developing new approaches for the recovery of missing data and lost information using the principle of signal sparsification, which includes compressed sensing and rank minimization. This research, funded by NSF, is enabling novel algorithms for image demosaicking (recovery of missing colors), deblurring, denoising, super-resolution, and visual coding.
New direction is being explored in the recovery of critical information embedded within massive amount of data and the reconstruction of high-resolution environmental data-fields from limited samples captured by sensor networks. This includes the application of more general principles of sparse coding for the recovery of multi-dimensional data fields. In particular, we are employing signal sparsification and rank-minimization principles for the recovery and reconstruction of high-resolution and high-dimension tensor data found in environmental modeling of a variety of data-fields. These data-fields could represent any physical or biological parameters that scientists and environmentalists are interested in.