scispace - formally typeset
Search or ask a question
Author

Joshin P. Krishnan

Other affiliations: Instituto Superior Técnico
Bio: Joshin P. Krishnan is an academic researcher from University of Agder. The author has contributed to research in topics: Topology (electrical circuits) & Computer science. The author has an hindex of 3, co-authored 11 publications receiving 30 citations. Previous affiliations of Joshin P. Krishnan include Instituto Superior Técnico.

Papers
More filters
Proceedings ArticleDOI
05 Jun 2021
TL;DR: In this article, the authors proposed an online kernel-based algorithm for topology estimation of non-linear vector autoregressive time series by solving a sparse online optimization framework using the composite objective mirror descent method.
Abstract: Estimating the unknown causal dependencies among graph-connected time series plays an important role in many applications, such as sensor network analysis, signal processing over cyber-physical systems, and finance engineering. Inference of such causal dependencies, often know as topology identification, is not well studied for non-linear non-stationary systems, and most of the existing methods are batch-based which are not capable of handling streaming sensor signals. In this paper, we propose an online kernel-based algorithm for topology estimation of non-linear vector autoregressive time series by solving a sparse online optimization framework using the composite objective mirror descent method. Experiments conducted on real and synthetic data sets show that the proposed algorithm outperforms the state-of-the-art methods for topology estimation.

15 citations

Journal ArticleDOI
16 Nov 2018-Sensors
TL;DR: In this paper, a dictionary learning phase retrieval (DLPR) algorithm is proposed to jointly learn the referred to dictionary and reconstructs the unknown target image by learning a complex domain dictionary from the data it represents via matrix factorization.
Abstract: This paper proposes a novel algorithm for image phase retrieval, i.e., for recovering complex-valued images from the amplitudes of noisy linear combinations (often the Fourier transform) of the sought complex images. The algorithm is developed using the alternating projection framework and is aimed to obtain high performance for heavily noisy (Poissonian or Gaussian) observations. The estimation of the target images is reformulated as a sparse regression, often termed sparse coding, in the complex domain. This is accomplished by learning a complex domain dictionary from the data it represents via matrix factorization with sparsity constraints on the code (i.e., the regression coefficients). Our algorithm, termed dictionary learning phase retrieval (DLPR), jointly learns the referred to dictionary and reconstructs the unknown target image. The effectiveness of DLPR is illustrated through experiments conducted on complex images, simulated and real, where it shows noticeable advantages over the state-of-the-art competitors.

11 citations

Proceedings ArticleDOI
25 Oct 2021
TL;DR: In this article, a kernel-based algorithm for graph topology estimation is proposed, which uses a Fourier-based Random feature approximation to tackle the curse of dimensionality associated with the kernel representations.
Abstract: Online topology estimation of graph-connected time series is challenging, especially since the causal dependencies in many real-world networks are nonlinear. In this paper, we propose a kernel-based algorithm for graph topology estimation. The algorithm uses a Fourier-based Random feature approximation to tackle the curse of dimensionality associated with the kernel representations. Exploiting the fact that the real-world networks often exhibit sparse topologies, we propose a group lasso based optimization framework, which is solve using an iterative composite objective mirror descent method, yielding an online algorithm with fixed computational complexity per iteration. The experiments conducted on real and synthetic data show that the proposed method outperforms its competitors.

6 citations

Posted Content
TL;DR: The algorithm, termed dictionary learning phase retrieval (DLPR), jointly learns the referred to dictionary and reconstructs the unknown target image and shows noticeable advantages over the state-of-the-art competitors.
Abstract: This paper proposes a novel algorithm for image phase retrieval, i.e., for recovering complex-valued images from the amplitudes of noisy linear combinations (often the Fourier transform) of the sought complex images. The algorithm is developed using the alternating projection framework and is aimed to obtain high performance for heavily noisy (Poissonian or Gaussian) observations. The estimation of the target images is reformulated as a sparse regression, often termed sparse coding, in the complex domain. This is accomplished by learning a complex domain dictionary from the data it represents via matrix factorization with sparsity constraints on the code (i.e., the regression coefficients). Our algorithm, termed dictionary learning phase retrieval (DLPR), jointly learns the referred to dictionary and reconstructs the unknown target image. The effectiveness of DLPR is illustrated through experiments conducted on complex images, simulated and real, where it shows noticeable advantages over the state-of-the-art competitors.

5 citations

Posted Content
TL;DR: In this paper, the authors proposed an online kernel-based algorithm for topology estimation of non-linear vector autoregressive time series by solving a sparse online optimization framework using the composite objective mirror descent method.
Abstract: Estimating the unknown causal dependencies among graph-connected time series plays an important role in many applications, such as sensor network analysis, signal processing over cyber-physical systems, and finance engineering. Inference of such causal dependencies, often know as topology identification, is not well studied for non-linear non-stationary systems, and most of the existing methods are batch-based which are not capable of handling streaming sensor signals. In this paper, we propose an online kernel-based algorithm for topology estimation of non-linear vector autoregressive time series by solving a sparse online optimization framework using the composite objective mirror descent method. Experiments conducted on real and synthetic data sets show that the proposed algorithm outperforms the state-of-the-art methods for topology estimation.

4 citations


Cited by
More filters
01 Jan 2016
TL;DR: Two-dimensional phase unwrapping algorithms applied to feminist theory crime and social justice theoretical conscience volume 4 dr-caloriez henry and the paper route cafebr chapter 3 what is money mishkin cafebr.
Abstract: two–dimensional phase unwrapping. theory, algorithms, and two dimensional phase unwrapping theory algorithms and two dimensional phase unwrapping theory algorithms and two-dimensional phase unwrapping using neural networks two-dimensional phase unwrapping: theory, algorithms, and (size 43,32mb) link download two dimensional phase phase unwrapping: project liverpool john moores university pixel-wise absolute phase unwrapping using geometric 2d phase unwrapping on fpgas and gpus phase unwrapping producing bright bands if phase unwrapping and affine transformations using cuda phase unwrapping on reconfigurable hardware ll.mit absolute three-dimensional shape measurement using coded fast twodimensional simultaneous phase unwrapping and low unwrapping differential x-ray phase-contrast images connections between transport of intensity equation and space geodesy seminar sio 239 scripps institution of experiment of phase unwrapping algorithm in interferometric reference documents esa 3d shape measurement technique for multiple rapidly moving phase unwrapping for large sar interferograms: statistical superfast phaseshifting method for 3-d shape measurement space geodesy seminar sio 239 scripps institution of off-axis quantitative phase imaging processing using cuda angular phase unwrapping of optically thick objects with a a comparison of phase unwrapping techniques in synthetic noise robust linear dynamic system for phase unwrapping fast phase processing in off-axis holography by cuda cat d2 dozer manual fiores fourier analysis of rgb fringe-projection profilometry and dynamic quantitative phase imaging for biological objects twowavelength quantitative phase unwrapping of dynamic comparison of phase unwrapping algorithms applied to feminist theory crime and social justice theoretical conscience volume 4 dr-caloriez henry and the paper route cafebr chapter 3 what is money mishkin cafebr

509 citations

Book ChapterDOI
12 Mar 2012

185 citations

Journal Article
Zhang Fe1
TL;DR: The results show that the PhaseCut method can obtain better reconstruction results than the traditional method which is based on binary masks, and the recovered phase distributions are loaded to the optical reconstruction system based on liquid crystal on silicon(LCOS).
Abstract: The classical iterative phase retrieval method has the disadvantages of poor anti-noise performance,and can not converge to the global optimal solution.The optimization of phase retrieval technology via PhaseCut(PC)transforms the phase retrieval problem into quadratic constrained convex programming problem,and gets the global optimal solution up to global phase.Some more structured and less random ternary masks and octonary masks are used to code the signal of interest to acquire diffraction patterns which can recover the missing phase information by solving PhaseCut.Simulations are firstly performed to test the one dimensional complex signal and compare the success rates of reconstruction using different masks.Secondly,the reconstruction results of two dimensional molecular images are compared for different masks.The results show that the method can obtain better reconstruction results than the traditional method which is based on binary masks.Moreover,the recovered phase distributions are loaded to the optical reconstruction system based on liquid crystal on silicon(LCOS).The diffraction pattern of real experiment proves the effectiveness of the proposed method.

77 citations

Journal ArticleDOI
19 Mar 2020
TL;DR: A deep-learning-based algorithm dedicated to the processing of speckle noise in phase measurements in digital holographic interferometry is presented, and the proposed approach exhibits state-of-the-art results.
Abstract: This paper presents a deep-learning-based algorithm dedicated to the processing of speckle noise in phase measurements in digital holographic interferometry. The deep learning architecture is trained with phase fringe patterns including faithful speckle noise, having non-Gaussian statistics and non-stationary property, and exhibiting spatial correlation length. The performances of the speckle de-noiser are estimated with metrics, and the proposed approach exhibits state-of-the-art results. In order to train the network to de-noise phase fringe patterns, a database is constituted with a set of noise-free and speckled phase data. The algorithm is applied to de-noising experimental data from wide-field digital holographic vibrometry. Comparison with the state-of-the-art algorithm confirms the achieved performance.

29 citations

Journal ArticleDOI
TL;DR: In this paper, the authors have studied phase unwrapping algorithms based on solving the discrete Poisson equation and compared their performance in terms of accuracy and efficiency, and an iteration strategy was introduced and its performance was investigated under different noise conditions.
Abstract: Phase unwrapping is a crucial process to obtain the absolute phase profile in many optical phase measurement techniques such as interferometry, holography, profilometry, etc. In this paper, we have studied several phase unwrapping algorithms based on solving the discrete Poisson equation. The differences among those algorithms lie in two aspects: one is the way to calculate the input for the Poisson equation using the wrapped phase data and the other is the way to compute the output (unwrapped phase data) using the corresponding input. Firstly, the method to compute the input for the Poisson equation was investigated using the finite difference (FD) and fast Fourier transform(FFT) method. Then different methods, based on FFT or discrete cosine transform (DCT), were employed to calculate the unwrapped phase, and their performances were compared in terms of accuracy and efficiency. To enhance the precision of those algorithms, an iteration strategy was introduced and its performance was investigated under different noise conditions. Finally, several real phase data were tested by using the direct and iterative methods. The detailed software package can be found in https://www.mathworks.com/matlabcentral/fileexchange/71810-phase- unwrapping-algorithms-by-solving-the-poisson-equation.

16 citations