Systems | Information | Learning | Optimization

Sparse Inverse Problems: Theory, Algorithms, and Applications

A common prior in modern statistical signal processing is that an observed signal is the noisy measurement of a few weighted sources. For example, the image of a collection of point sources of light can be parameterized by their locations and intensities. The goal of superresolution is to remove the blur induced by diffraction by solving for the locations and intensities of the individual point sources. This talk will introduce the basics of sparse inverse problems, present an algorithm to solve them, and discuss some theoretical guarantees for recovery. Throughout, we will focus on superresolution imaging as an example of a sparse inverse problem.
September 23 @ 12:30
12:30 pm (1h)

Discovery Building, Orchard View Room

Geoffrey Schiebinger