The Newton method for affine phase retrieval

Abstract. We consider the problem of recovering a signal from the magnitudes of affine measurements, which is also known as affine phase retrieval. In this paper, we formulate affine phase retrieval as an optimization problem and develop a second-order algorithm based on Newton method to solve it. Besides being able to convert into a phase retrieval problem, affine phase retrieval has its unique advantages in its solution. For example, the linear information in the observation makes it possible to solve this problem with second-order algorithms under complex measurements. Another advantage is that our algorithm doesn’t have any special requirements for the initial point, while an appropriate initial value is essential for most non-convex phase retrieval algorithms. Starting from zero, our algorithm generates iteration point by Newton method, and we prove that the algorithm can quadratically converge to the true signal without any ambiguity for both Gaussian measurements and CDP measurements. In addition, we also use some numerical simulations to verify the conclusions and to show the effectiveness of the algorithm.

[1]  Yonina C. Eldar,et al.  Solving Systems of Random Quadratic Equations via Truncated Amplitude Flow , 2016, IEEE Transactions on Information Theory.

[2]  Chinmay Hegde,et al.  Fast, Sample-Efficient Algorithms for Structured Phase Retrieval , 2017, NIPS.

[3]  J. Miao,et al.  Extending X-ray crystallography to allow the imaging of noncrystalline materials, cells, and single protein complexes. , 2008, Annual review of physical chemistry.

[4]  Yonina C. Eldar,et al.  STFT Phase Retrieval: Uniqueness Guarantees and Recovery Algorithms , 2015, IEEE Journal of Selected Topics in Signal Processing.

[5]  Ziyang Yuan,et al.  Phase retrieval with background information , 2018, Inverse Problems.

[6]  Yonina C. Eldar,et al.  Fourier Phase Retrieval: Uniqueness and Algorithms , 2017, ArXiv.

[7]  M. Salman Asif,et al.  Fourier Phase Retrieval with Side Information Using Generative Prior , 2019, 2019 53rd Asilomar Conference on Signals, Systems, and Computers.

[8]  Veit Elser,et al.  Benchmark problems for phase retrieval , 2017, SIAM J. Imaging Sci..

[9]  Yuxin Chen,et al.  Solving Random Quadratic Systems of Equations Is Nearly as Easy as Solving Linear Systems , 2015, NIPS.

[10]  Rick P. Millane,et al.  Phase retrieval in crystallography and optics , 1990 .

[11]  M. Salman Asif,et al.  Solving Phase Retrieval with a Learned Reference , 2020, ECCV.

[12]  Bing Gao,et al.  Phaseless Recovery Using the Gauss–Newton Method , 2016, IEEE Transactions on Signal Processing.

[13]  Xiaodong Li,et al.  Phase Retrieval via Wirtinger Flow: Theory and Algorithms , 2014, IEEE Transactions on Information Theory.

[14]  Bing Gao,et al.  Phase Retrieval From the Magnitudes of Affine Linear Measurements , 2016, Adv. Appl. Math..

[15]  Thierry Blu,et al.  Local amplitude and phase retrieval method for digital holography applied to microscopy , 2003, European Conference on Biomedical Optics.

[16]  Xiaodong Li,et al.  Phase Retrieval from Coded Diffraction Patterns , 2013, 1310.3240.

[17]  Prateek Jain,et al.  Phase Retrieval Using Alternating Minimization , 2013, IEEE Transactions on Signal Processing.

[18]  O. Bunk,et al.  Diffractive imaging for periodic samples: retrieving one-dimensional concentration profiles across microfluidic channels. , 2007, Acta crystallographica. Section A, Foundations of crystallography.

[19]  A. Walther.  The Question of Phase Retrieval in Optics , 1963 .

[20]  Babak Hassibi,et al.  Sparse Phase Retrieval: Uniqueness Guarantees and Recovery Algorithms , 2013, IEEE Transactions on Signal Processing.

[21]  M. Salman Asif,et al.  Fourier Phase Retrieval with Arbitrary Reference Signal , 2020, ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).