Real-time phase-retrieval and wavefront sensing enabled by an artificial neural network

White, James; Wang, Sici; Eschen, Wilhelm; Rothhardt, Jan

doi:10.1364/oe.419105

Cited by 8 publications

(6 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Third, there is the rising field of machine learning [ 22 ] . There are vastly different approaches here, but many of them include the usage of phase masks that generate an extended diffraction pattern in the FF in order to collect more data about the phase [ 23 ] , or even implement diffractive neural networks in front of the FF sensor [ 24 ] . Then, a model is trained on a large collection of known data points in order to be able to make single-shot estimations of the NF phase.…”

Section: Discussion and Outlookmentioning

confidence: 99%

“…Then, a model is trained on a large collection of known data points in order to be able to make single-shot estimations of the NF phase. These approaches have been proven to be versatile, even granting insight into STCs, and provide fast phase estimations once the training has been successfully completed [ 23 ] . However, the generation of such a model is challenging in many aspects, including the loss of intuitive understanding about the beam and the recording of a suitable and sufficiently large dataset for training and verification.…”

Section: Discussion and Outlookmentioning

confidence: 99%

See 1 more Smart Citation

Zernike-coefficient extraction via helical beam reconstruction for optimization (ZEHBRO) in the far field

Ohland,

Posor,

Eisenbarth

et al. 2023

High Pow Laser Sci Eng

View full text Add to dashboard Cite

The spatial distribution of beams with orbital angular momentum in the far field is known to be extremely sensitive to angular aberrations, such as astigmatism, coma and trefoil. This poses a challenge for conventional beam optimization strategies when a homogeneous ring intensity is required for an application. We developed a novel approach for estimating the Zernike coefficients of low-order angular aberrations in the near field based solely on the analysis of the ring deformations in the far field. A fast, iterative reconstruction of the focal ring recreates the deformations and provides insight into the wavefront deformations in the near field without relying on conventional phase retrieval approaches. The output of our algorithm can be used to optimize the focal ring, as demonstrated experimentally at the 100 TW beamline at the Extreme Light Infrastructure -Nuclear Physics facility.

show abstract

Section: Discussion and Outlookmentioning

confidence: 99%

Section: Discussion and Outlookmentioning

confidence: 99%

Zernike-coefficient extraction via helical beam reconstruction for optimization (ZEHBRO) in the far field

Ohland,

Posor,

Eisenbarth

et al. 2023

High Pow Laser Sci Eng

View full text Add to dashboard Cite

show abstract

“…After each excitation, the next task is to retrieve the corresponding wavefront of the output beams. There are many possible approaches to tackle this problem, such as coherent diffractive imaging techniques [5] and phase retrieval neural networks [6]. In this work, the traditional Gerchberg-Saxton (GS) algorithm [7] was employed, for which two beam profiles along the propagation path of the output beam are needed.…”

Section: Methodsmentioning

confidence: 99%

Characterizing the transverse modes of optical fibers by singular value decomposition

Tu¹,

Pfleghar²,

Jáuregui³

et al. 2023

Fiber Lasers XX: Technology and Systems

View full text Add to dashboard Cite

A new approach to experimentally characterize the transverse modes of optical fibers is proposed in this submission. It analyzes a large volume of electric field data captured from the fiber under test and obtains the orthogonal modal base set of the fiber using the singular value decomposition. This procedure is similar to the principle of machine learning in the area of artificial intelligence. The results show a good agreement with the simulated transverse modes. Due to its operating principle, this approach can characterize any fiber regardless of its length and size.

show abstract

“…Instead of solving the phase problem by alternating projection, machine learning (ML) based methods also show the potential for solving inverse problems [21] [22]. ML model has been considered as universal approximators that can adequate to complex and non-linear computational task [23] [24]. Unlike most applications that learn concealed pattern features by large sample training data, the ML solution for inverse problems concentrates more on the integration with the physical model of the experimental setup to maximize the consequence of all the information available.…”

Section: Introductionmentioning

confidence: 99%

Rapid profile reconstruction of phase defects via machine-learning regression model

He¹,

Tan²,

et al. 2022

International Conference on Extreme Ultraviolet Lithography 2022

View full text Add to dashboard Cite

For imaging methods based on Fourier transform, such as coherent diffraction imaging (CDI), in the case where the sampling rate of the diffraction data satisfies Nyquist sampling theorem, phase retrieval algorithm can effectively recover the phase lost in the image acquisition process. In general, ptychography is used to ensure the adequate sample. But in the imaging system of EUV mask defect inspection, the acquisition of more sampling data requires multiple exposures of the mask, which increases the sampling time resulting in low throughput and causes mask irradiation damage. To solve the problem, we propose a machine learning scheme to reconstruct the profile of phase defect using a single diffraction intensity. Combining the physical model of CDI and neural network framework, the method transforms the phase retrieval problem into a regression model of machine learning. Integrating the constraints of the traditional CDI and the gradient descent algorithms in neural network framework, our approach can accelerate the convergence of the model. The performance of the method is confirmed by comparing with the conventional iterative method in the condition of sufficient sampling rate. This novel method provides a new idea of combining computational imaging with machine learning, which can simplify the data acquisition process and accelerate the iteration of algorithm in EUV mask defect inspection.

show abstract

Real-time phase-retrieval and wavefront sensing enabled by an artificial neural network

Cited by 8 publications

References 14 publications

Zernike-coefficient extraction via helical beam reconstruction for optimization (ZEHBRO) in the far field

Zernike-coefficient extraction via helical beam reconstruction for optimization (ZEHBRO) in the far field

Characterizing the transverse modes of optical fibers by singular value decomposition

Rapid profile reconstruction of phase defects via machine-learning regression model

Contact Info

Product

Resources

About