arXiv subjects

Daniel S. Weller

Publications and source records attributed to Daniel S. Weller.

8 recordsLinked to original sources

arXiv2020

PrecoG: an efficient unitary split preconditioner for the transform-domain LMS filter via graph Laplacian regularization

Transform-domain least mean squares (LMS) adaptive filters encompass the class of algorithms where the input data are subjected to a data-independent unitary transform followed by a power normalization stage as preprocessing steps. Because conventional transformations are not data-dependent, this preconditioning procedure was shown theoretically to improve the convergence of the LMS filter only for certain classes of input data. However, in reality if the class of input data is not known beforehand, it is difficult to decide which transformation to use. Thus, there is a need to devise a learning framework to obtain such a preconditioning transformation using input data prior to applying on the input data. It is hypothesized that the underlying topology of the data affects the selection of the transformation. With the input modeled as a weighted graph that mimics neuronal interactions, PrecoG obtains the desired transform by recursive estimation of the graph Laplacian matrix. Additionally, we show the efficacy of the transform as a generalized split preconditioner on a linear system of equations and in Hebb-LMS settings. In terms of the improvement of the condition number after applying the transformation, PrecoG performs significantly better than the existing state-of-the-art techniques that involve unitary and non-unitary transforms.

eess.SP↗

arXiv2019

Enhanced Center Coding for Cell Detection with Convolutional Neural Networks

Cell imaging and analysis are fundamental to biomedical research because cells are the basic functional units of life. Among different cell-related analysis, cell counting and detection are widely used. In this paper, we focus on one common step of learning-based cell counting approaches: coding the raw dot labels into more suitable maps for learning. Two criteria of coding raw dot labels are discussed, and a new coding scheme is proposed in this paper. The two criteria measure how easy it is to train the model with a coding scheme, and how robust the recovered raw dot labels are when predicting. The most compelling advantage of the proposed coding scheme is the ability to distinguish neighboring cells in crowded regions. Cell counting and detection experiments are conducted for five coding schemes on four types of cells and two network architectures. The proposed coding scheme improves the counting accuracy versus the widely-used Gaussian and rectangle kernels up to 12%, and also improves the detection accuracy versus the common proximity coding up to 14%.

cs.CV↗

arXiv2019

Hieroglyph: Hierarchical Glia Graph Skeletonization and Matching

Automatic 3D reconstruction of glia morphology is a powerful tool necessary for investigating the role of microglia in neurological disorders in the central nervous system. Current glia skeleton reconstruction techniques fail to capture an accurate tracing of the processes over time, useful for the study of the microglia motility and morphology in the brain during healthy and diseased states. We propose Hieroglyph, a fully automatic temporal 3D skeleton reconstruction algorithm for glia imaged via 3D multiphoton microscopy. Hieroglyph yielded a 21% performance increase compared to state of the art automatic skeleton reconstruction methods and outperforms the state of the art in different measures of consistency on datasets of 3D images of microglia. The results from this method provide a 3D graph and digital reconstruction of glia useful for a myriad of morphological analyses that could impact studies in brain immunology and disease.

eess.IV↗

arXiv2016

Comparison-based Image Quality Assessment for Parameter Selection

Image quality assessment (IQA) is traditionally classified into full-reference (FR) IQA and no-reference (NR) IQA according to whether the original image is required. Although NR-IQA is widely used in practical applications, room for improvement still remains because of the lack of the reference image. Inspired by the fact that in many applications, such as parameter selection, a series of distorted images are available, the authors propose a novel comparison-based image quality assessment (C-IQA) method. The new comparison-based framework parallels FR-IQA by requiring two input images, and resembles NR-IQA by not using the original image. As a result, the new comparison-based approach has more application scenarios than FR-IQA does, and takes greater advantage of the accessible information than the traditional single-input NR-IQA does. Further, C-IQA is compared with other state-of-the-art NR-IQA methods on two widely used IQA databases. Experimental results show that C-IQA outperforms the other NR-IQA methods for parameter selection, and the parameter trimming framework combined with C-IQA saves the computation of iterative image reconstruction up to 80%.

cs.CV↗

arXiv2014

Undersampled Phase Retrieval with Outliers

We propose a general framework for reconstructing transform-sparse images from undersampled (squared)-magnitude data corrupted with outliers. This framework is implemented using a multi-layered approach, combining multiple initializations (to address the nonconvexity of the phase retrieval problem), repeated minimization of a convex majorizer (surrogate for a nonconvex objective function), and iterative optimization using the alternating directions method of multipliers. Exploiting the generality of this framework, we investigate using a Laplace measurement noise model better adapted to outliers present in the data than the conventional Gaussian noise model. Using simulations, we explore the sensitivity of the method to both the regularization and penalty parameters. We include 1D Monte Carlo and 2D image reconstruction comparisons with alternative phase retrieval algorithms. The results suggest the proposed method, with the Laplace noise model, both increases the likelihood of correct support recovery and reduces the mean squared error from measurements containing outliers. We also describe exciting extensions made possible by the generality of the proposed framework, including regularization using analysis-form sparsity priors that are incompatible with many existing approaches.

stat.AP↗

arXiv2010

Bayesian Post-Processing Methods for Jitter Mitigation in Sampling

Minimum mean squared error (MMSE) estimators of signals from samples corrupted by jitter (timing noise) and additive noise are nonlinear, even when the signal prior and additive noise have normal distributions. This paper develops a stochastic algorithm based on Gibbs sampling and slice sampling to approximate the optimal MMSE estimator in this Bayesian formulation. Simulations demonstrate that this nonlinear algorithm can improve significantly upon the linear MMSE estimator, as well as the EM algorithm approximation to the maximum likelihood (ML) estimator used in classical estimation. Effective off-chip post-processing to mitigate jitter enables greater jitter to be tolerated, potentially reducing on-chip ADC power consumption.

stat.AP↗

arXiv2010

On the Estimation of Nonrandom Signal Coefficients from Jittered Samples

This paper examines the problem of estimating the parameters of a bandlimited signal from samples corrupted by random jitter (timing noise) and additive iid Gaussian noise, where the signal lies in the span of a finite basis. For the presented classical estimation problem, the Cramer-Rao lower bound (CRB) is computed, and an Expectation-Maximization (EM) algorithm approximating the maximum likelihood (ML) estimator is developed. Simulations are performed to study the convergence properties of the EM algorithm and compare the performance both against the CRB and a basic linear estimator. These simulations demonstrate that by post-processing the jittered samples with the proposed EM algorithm, greater jitter can be tolerated, potentially reducing on-chip ADC power consumption substantially.

stat.AP↗

arXiv2008

Nonlinear Digital Post-Processing to Mitigate Jitter in Sampling

This paper describes several new algorithms for estimating the parameters of a periodic bandlimited signal from samples corrupted by jitter (timing noise) and additive noise. Both classical (non-random) and Bayesian formulations are considered: an Expectation-Maximization (EM) algorithm is developed to compute the maximum likelihood (ML) estimator for the classical estimation framework, and two Gibbs samplers are proposed to approximate the Bayes least squares (BLS) estimate for parameters independently distributed according to a uniform prior. Simulations are performed to demonstrate the significant performance improvement achievable using these algorithms as compared to linear estimators. The ML estimator is also compared to the Cramer-Rao lower bound to determine the range of jitter for which the estimator is approximately efficient. These simulations provide evidence that the nonlinear algorithms derived here can tolerate 1.4-2 times more jitter than linear estimators, reducing on-chip ADC power consumption by 50-75 percent.

stat.AP↗