作者：
Desmal, AbdullaBagci, HakanKAUST
Div Comp Elect & Math Sci & Engn Thuwal Saudi Arabia KAUST
Ctr Uncertainty Quantificat Computat Sci & Engn Thuwal Saudi Arabia
A numerical framework that incorporates recently developed iterativeshrinkagethresholding (IST) algorithms within the Born iterative method (BIM) is proposed for solving the two-dimensional inverse electromagnetic s...
详细信息
A numerical framework that incorporates recently developed iterativeshrinkagethresholding (IST) algorithms within the Born iterative method (BIM) is proposed for solving the two-dimensional inverse electromagnetic scattering problem. IST algorithms minimize a cost function weighted between measurement-data misfit and a zeroth/first-norm penalty term and therefore promote "sharpness" in the solution. Consequently, when applied to domains with sharp variations, discontinuities, or sparse content, the proposed framework is more efficient and accurate than the "classical" BIM that minimizes a cost function with a second-norm penalty term. Indeed, numerical results demonstrate the superiority of the IST-BIM over the classical BIM when they are applied to sparse domains: Permittivity and conductivity profiles recovered using the IST-BIM are sharper and more accurate and converge faster.
Computer-assisted medical diagnosis has grown its popularity among researchers and practitioners due to its high applicability and cost-effective applications. The use of learning methods in facilitating the processin...
详细信息
ISBN:
(纸本)9789811992247;9789811992254
Computer-assisted medical diagnosis has grown its popularity among researchers and practitioners due to its high applicability and cost-effective applications. The use of learning methods in facilitating the processing and analyses of Medical Images has become necessary. Image de-blurring problem in Medical Imaging applications is a challenging task to improve the quality of the images. The class of iterative shrinkage thresholding algorithms (ISTA) is considered for handling the image de-blurring problems. These algorithms use first-order information and are tempting due to their simplicity. However, they converge pretty slowly. In this work, first, a faster version of ISTA is proposed by utilizing Nestrove's updating rule. Later, the proposed approach is used to solve the image de-blurring problems arising in Medical Imaging. A theoretical study is performed to verify the improvements in the convergence rate of the modified fast ISTA. Numerical experiments suggest that the presented method for the image de-blurring problem outperforms the baseline ISTA method.
Convolutional sparse coding model has been successfully used in some tasks such as signal or image processing and classification. The recently proposed supervised convolutional sparse coding network (CSCNet) model bas...
详细信息
ISBN:
(纸本)9783031263187;9783031263194
Convolutional sparse coding model has been successfully used in some tasks such as signal or image processing and classification. The recently proposed supervised convolutional sparse coding network (CSCNet) model based on the Minimum Mean Square Error (MMSE) approximation shows the similar PSNR value for image denoising problem with state of the art methods while using much fewer parameters. The CSCNet uses the learning convolutional iterativeshrinkage-thresholdingalgorithms (LISTA) based on the convolutional dictionary setting. However, LISTA methods are known to converge to local minima. In this paper we proposed one novel algorithm based on LISTA with dry friction, named LISTDFA. The dry friction enters the LISTDFA algorithm through proximal mapping. Due to the nature of dry friction, the LISTDFA algorithm is proven to converge in a finite time. The corresponding iterative neural network preserves the computational simplicity of the original CSCNet, and can reach a better local minima practically.
暂无评论