Blind Deconvolution Meets Blind Demixing: Algorithms and Performance Bounds by Shuyang Ling, Thomas Strohmer
Image Credit: NASA/JPL-Caltech
This image was taken by Rear Hazcam: Right B (RHAZ_RIGHT_B) onboard NASA's Mars rover Curiosity on Sol 1219 (2016-01-10 14:29:24 UTC).
Full Resolution
Consider $r$ sensors, each one intends to send a function $x_i$ (e.g. a signal or image) to a receiver common to all $r$ sensors. Before transmission, each $x_i$ is multiplied by an "encoding matrix" $A_i$. During transmission each $A_ix_i$ gets convolved with a function $h_i$. The receiver records the function $y$, given by the sum of all these convolved signals. Assume that the receiver knowns all the $A_i$, but does neither know the $x_i$ nor the $h_i$. When and under which conditions is it possible to recover the individual signals $x_i$ and the channels $h_i$ from just one received signal $y$? This challenging problem, which intertwines blind deconvolution with blind demixing, appears in a variety of applications, such as audio processing, image processing, neuroscience, spectroscopy, and astronomy. It is also expected to play a central role in connection with the future Internet-of-Things. We will prove that under reasonable and practical assumptions, it is possible to solve this otherwise highly ill-posed problem and recover the $r$ transmitted functions $x_i$ and the impulse responses $h_i$ in a robust, reliable, and efficient manner from just one single received function $y$ by solving a semidefinite program. We derive explicit bounds on the number of measurements needed for successful recovery and prove that our method is robust in presence of noise. Our theory is actually a bit pessimistic, since numerical experiments demonstrate that, quite remarkably, recovery is still possible if the number of measurements is close to the number of degrees of freedom.
Image Credit: NASA/JPL-Caltech
This image was taken by Rear Hazcam: Right B (RHAZ_RIGHT_B) onboard NASA's Mars rover Curiosity on Sol 1219 (2016-01-10 14:29:24 UTC).
Full Resolution
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.
No comments:
Post a Comment