Signal recovery via deep convolutional networks

Date
2021-04-20
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract

Real-world data may not be sparse in a fixed basis, and current high-performance recovery algorithms are slow to converge, which limits compressive sensing (CS) to either non-real-time applications or scenarios where massive back-end computing is available. Presented herein are embodiments for improving CS by developing a new signal recovery framework that uses a deep convolutional neural network (CNN) to learn the inverse transformation from measurement signals. When trained on a set of representative images, the network learns both a representation for the signals and an inverse map approximating a greedy or convex recovery algorithm. Implementations on real data indicate that some embodiments closely approximate the solution produced by state-of-the-art CS recovery algorithms, yet are hundreds of times faster in run time.

Description
Advisor
Degree
Type
Utility patent
Keywords
Citation

Baraniuk, Richard G. and Mousavi, Ali, "Signal recovery via deep convolutional networks." Patent US10985777B2. issued 2021-04-20. Retrieved from https://hdl.handle.net/1911/110644.

Has part(s)
Forms part of
Published Version
Rights
Link to license
Citable link to this page
Collections