Wide-field wavefront sensing with convolutional neural networks and ordinary least squares

2020 ◽  
Author(s):  
David Thomas ◽  
Joshua E. Meyers ◽  
Steven M. Kahn
Keyword(s):  
Neural Networks ◽  
Least Squares ◽  
Convolutional Neural Networks ◽  
Ordinary Least Squares ◽  
Wavefront Sensing ◽  
Wide Field

Background Rejection using Convolutional Neural Networks

2017 ◽  
Vol 13 (S338) ◽  
pp. 37-39
Author(s):  
Adam Zadrożny ◽  
Beata Goźlińska
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Early Stage ◽  
Simulated Data ◽  
Wide Field ◽  
Proof Of Concept ◽  
Field Surveys ◽  
Background Rejection ◽  
Very High

AbstractThe paper presents a proof of concept method of background rejection based on convolutional neural networks (CNN). The method was tested on simulated data and achieved very high accuracy (100%). What is more, method based on CNN is very fast and could be easily applied to wide field surveys. Since early stage results suggest method is very accurate and robust, it could be helpful in creating very low-latency pipelines for EM Follow-up purposes, which will be needed in LIGO-Virgo O3 EM Follow-up.

scholarly journals A weighted least squares optimisation strategy for medical image super resolution via multiscale convolutional neural networks for healthcare applications

2021 ◽  
Author(s):  
Bhawna Goyal ◽  
Dawa Chyophel Lepcha ◽  
Ayush Dogra ◽  
Shui-Hua Wang
Keyword(s):  
Neural Networks ◽  
Least Squares ◽  
Convolutional Neural Networks ◽  
Medical Image ◽  
Medical Images ◽  
Weighted Least Squares ◽  
Super Resolution ◽  
Poor Quality ◽  
Healthcare Applications ◽  
Image Super Resolution

AbstractMedical imaging is an essential medical diagnosis system subsequently integrated with artificial intelligence for assistance in clinical diagnosis. The actual medical images acquired during the image capturing procedures generate poor quality images as a result of numerous physical restrictions of the imaging equipment and time constraints. Recently, medical image super-resolution (SR) has emerged as an indispensable research subject in the community of image processing to address such limitations. SR is a classical computer vision operation that attempts to restore a visually sharp high-resolution images from the degraded low-resolution images. In this study, an effective medical super-resolution approach based on weighted least squares optimisation via multiscale convolutional neural networks (CNNs) has been proposed for lesion localisation. The weighted least squares optimisation strategy that particularly is well-suited for progressively coarsening the original images and simultaneously extract multiscale information has been executed. Subsequently, a SR model by training CNNs based on wavelet analysis has been designed by carrying out wavelet decomposition of optimized images for multiscale representations. Then multiple CNNs have been trained separately to approximate the wavelet multiscale representations. The trained multiple convolutional neural networks characterize medical images in many directions and multiscale frequency bands, and thus facilitate image restoration subject to increased number of variations depicted in different dimensions and orientations. Finally, the trained CNNs regress wavelet multiscale representations from a LR medical images, followed by wavelet synthesis that forms a reconstructed HR medical image. The experimental performance indicates that the proposed model SR restoration approach achieve superior SR efficiency over existing comparative methods

Least-squares reverse time migration using convolutional neural networks

Geophysics ◽  
2021 ◽  
pp. 1-92
Author(s):  
Wei Zhang ◽  
Jinghuai Gao ◽  
Tao Yang ◽  
Xiudi Jiang ◽  
Wenbo Sun
Keyword(s):  
Neural Networks ◽  
High Resolution ◽  
Least Squares ◽  
Convolutional Neural Networks ◽  
Input Data ◽  
Velocity Model ◽  
Reverse Time ◽  
Reverse Time Migration ◽  
Time Migration ◽  
Reflection Image

Least-squares reverse time migration (LSRTM) has the potential to reconstruct a high-resolution image of subsurface reflectivity. However, the current data-domain LSRTM approach, which iteratively updates the subsurface reflectivity by minimizing the data residuals, is a computationally expensive task. To alleviate this problem and improve imaging quality, we develop a LSRTM approach using convolutional neural networks (CNNs), which is referred to as CNN-LSRTM. Specifically, the LSRTM problem can be implemented via a gradient-like iterative scheme, in which the updating component in each iteration is learned via a CNN model. In order to make the most of observation data and migration velocity model at hand, we utilize the common-source RTM image, the stacked RTM image, and the migration velocity model rather than only the stacked RTM image as the input data of CNN. We have successfully trained the constructed CNN model on the training data sets with a total of 5000 randomly layered and fault models. Based on the well-trained CNN model, we have proved that the proposed approach can efficiently recover the high-resolution reflection image for the layered, fault, and overthrust models. Through a marine field data experiment, it can determine the benefit of our constructed CNN model in terms of computational efficiency. In addition, we analyze the influence of input data of the constructed CNN model on the reconstruction quality of the reflection image.

scholarly journals Application Of L1- Regularization Approach In QSAR Problem. Linear Regression And Artificial Neural Networks

2019 ◽  
Vol 14 (2) ◽  
pp. 79-90
Author(s):  
M.I. Berdnyk ◽  
A.B. Zakharov ◽  
V.V. Ivanov
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Least Squares ◽  
Principal Component Regression ◽  
Principal Component ◽  
Ordinary Least Squares ◽  
Regression Equations ◽  
Absolute Deviation ◽  
Internal Validation ◽  
Artificial Neural

One of the primary tasks of analytical chemistry and QSAR/QSPR researches is building of prognostic regression equations based on descriptors sets. The one of the most important problems here is to decrease the number of descriptors in the initial descriptor set which is usually way too big. In current investigation the descriptor set is proposed to be reduced employing the least absolute shrinkage and selection operator (LASSO) approach. Decreased descriptor sets were used for calculations with application of the following QSAR/QSPR methods: ordinary least squares (OLS), the least absolute deviation (LAD) regressions and artificial neural networks (ANN). Contrary to aforementioned methods principal component regression (PCR) and partial least squares (PLS) approaches can produce solutions containing numerous descriptors. In this article we compared the viability of these two different descriptor handling ideologies in application to molecular chemical and physical properties prediction. From the obtained results it is possible to see that there are tasks for which PCR and PLS approaches can fail to produce accurate regression equations. At the same time, methods OLS and LAD that use small amount of descriptors can provide viable solutions for the same cases. It was shown that these small sets of descriptors selected with LASSO approach can be used in ANN to obtain models with even better internal validation characteristics.

Convolutional neural networks for object-agnostic wavefront sensing in the presence of noise

2020 ◽  
Author(s):  
Kedar Naik ◽  
Raymond Wright ◽  
David Claveau ◽  
D S. Acton ◽  
J S. Knight
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Wavefront Sensing

scholarly journals Vetting the optical transient candidates detected by the GWAC network using convolutional neural networks

2020 ◽  
Vol 497 (3) ◽  
pp. 2641-2650
Author(s):  
Damien Turpin ◽  
M Ganet ◽  
S Antier ◽  
E Bertin ◽  
L P Xin ◽  
...  
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Large Fraction ◽  
Computing Time ◽  
Wide Field ◽  
Learning Approaches ◽  
Robotic Telescopes ◽  
Optical Domain ◽  
Automated Tools ◽  
Optical Transient

ABSTRACT The observation of the transient sky through a multitude of astrophysical messengers has led to several scientific breakthroughs in the last two decades, thanks to the fast evolution of the observational techniques and strategies employed by the astronomers. Now, it requires to be able to coordinate multiwavelength and multimessenger follow-up campaigns with instruments both in space and on ground jointly capable of scanning a large fraction of the sky with a high-imaging cadency and duty cycle. In the optical domain, the key challenge of the wide field-of-view telescopes covering tens to hundreds of square degrees is to deal with the detection, identification, and classification of hundreds to thousands of optical transient (OT) candidates every night in a reasonable amount of time. In the last decade, new automated tools based on machine learning approaches have been developed to perform those tasks with a low computing time and a high classification efficiency. In this paper, we present an efficient classification method using convolutional neural networks (CNNs) to discard many common types of bogus falsely detected in astrophysical images in the optical domain. We designed this tool to improve the performances of the OT detection pipeline of the Ground Wide field Angle Cameras (GWAC) telescopes, a network of robotic telescopes aiming at monitoring the OT sky down to R = 16 with a 15 s imaging cadency. We applied our trained CNN classifier on a sample of 1472 GWAC OT candidates detected by the real-time detection pipeline.

Sign in / Sign up

Export Citation Format

Share Document