Robust 1-bit compressive sensing using binary matching pursuit

Ming Yan let me know of a new preprint entitled Robust 1-bit compressive sensing using binary matching pursuit by Ming Yan, Yi Yang, Stanley Osher.  The abstract reads:

In compressive sensing (CS), the goal is to recover signals at reduced sample rate compared to the classic Shannon-Nyquist rate. However, the classic CS theory assumes the measurements to be real-valued and have finite bit precision. The quantization of CS measurements has been studied recently and it has been shown that accurate and stable signal acquisition is possible even when each measurement is quantized to only one single bit. There are many algorithms proposed for 1-bit compressive sensing and they work well when there is no noise in the measurements, e.g., there are no sign flips, while the performance is worsened when there are a lot of sign flips in the measurements. In this paper, we propose a robust method for recovering signals from 1-bit measurements using binary matching pursuit. This method will detect the positions where sign flips happen and recover the signals using \correct" measurements.Numerical experiments show the accuracy of sign flips detection and high performance of signal recovery for our algorithms compared with other algorithms.

The webpage for the project and the attendant implementation are available here at: 

http://www.math.ucla.edu/~yanm/BMP.html

Thanks Ming 

Maybe I should start a page dedicated to 1-bit compressive sensing ?

Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.