A conjugate gradient Wirtinger flow (CG-WF) algorithm for phase retrieval is proposed in this paper. It is shown that, compared with recently reported Wirtinger flow and its modified methods, the proposed CG-WF algorithm is able to dramatically accelerate the convergence rate while keeping the dominant computational cost of each iteration unchanged. We numerically illustrate the effectiveness of our method in recovering 1D Gaussian signals and 2D natural color images under both Gaussian and coded diffraction pattern models.
|Number of pages||5|
|Journal||Journal of the Optical Society of America A: Optics and Image Science, and Vision|
|Publication status||Published - 1 May 2017|
ASJC Scopus subject areas
- Electronic, Optical and Magnetic Materials
- Atomic and Molecular Physics, and Optics
- Computer Vision and Pattern Recognition