Scalable Plug-and-Play ADMM With Convergence Guarantees
Abstract
Plug-and-play priors (PnP) is a broadly applicable methodology for solving inverse problems by exploiting statistical priors specified as denoisers. Recent work has reported the state-of-the-art performance of PnP algorithms using pre-trained deep neural nets as denoisers in a number of imaging applications. However, current PnP algorithms are impractical in large-scale settings due to their heavy computational and memory requirements. This work addresses this issue by proposing an incremental variant of the widely used PnP-ADMM algorithm, making it scalable to problems involving a large number measurements. We theoretically analyze the convergence of the algorithm under a set of explicit assumptions, extending recent theoretical results in the area. Additionally, we show the effectiveness of our algorithm with nonsmooth data-fidelity terms and deep neural net priors, its fast convergence compared to existing PnP algorithms, and its scalability in terms of speed and memory.
Additional Information
© 2021 IEEE. Manuscript received January 14, 2021; revised April 27, 2021 and June 27, 2021; accepted June 27, 2021. Date of publication July 2, 2021; date of current version August 14, 2021. This work supported in part by the National Science Foundation award CCF-1813910, by the the National Science Foundation CAREER award under Grant CCF-2043134, and by the Laboratory Directed Research and Development program of Los Alamos National Laboratory under project number 20200061DR. The associate editor coordinating the review of this manuscript and approving it for publication was Prof. Stanley H. Chan. (Yu Sun, Zihui Wu, and Xiaojian Xu contributed equally to this work.)Attached Files
Submitted - 2006.03224.pdf
Files
Name | Size | Download all |
---|---|---|
md5:ad688727040fc4a77566c21735298ce3
|
3.0 MB | Preview Download |
Additional details
- Eprint ID
- 111155
- Resolver ID
- CaltechAUTHORS:20211001-191837956
- CCF-1813910
- NSF
- CCF-2043134
- NSF
- 20200061DR
- Los Alamos National Laboratory
- Created
-
2021-10-04Created from EPrint's datestamp field
- Updated
-
2021-10-04Created from EPrint's last_modified field