Higher order 1-bit Sigma-Delta modulation on a circle

The authors derive efficient parallel algorithms on a VLSI array for several signal and image processing tasks. The parallel architecture has n/sup 2/ memory modules storing the input data which is accessed by n PEs. They consider several signal processing tasks such as k selection, median filtering, labeling 0/1 image, etc. For these problems, linear speedup is obtained, compared to a single-processor system. For the k selection problem they also provide an efficient parallel solution using log transformation on the data. The expected time of this method is O(n) with a small constant factor. The proposed array is suitable for general-purpose signal processing and can be implemented in VLSI using a limited chip set.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Models of signal processing in human hearing

Representing data residing on a graph as a linear combination of building block signals can enable efficient and insightful visual or statistical analysis of the data, and such representations prove useful as regularizers in signal processing and machine learning tasks. Designing collections of building block signals -- or more formally, dictionaries of atoms -- that specifically account for the underlying graph structure as well as any available representative training signals has been an active area of research over the last decade. In this article, we survey a particular class of dictionaries called localized spectral graph filter frames, whose atoms are created by localizing spectral patterns to different regions of the graph. After showing how this class encompasses a variety of approaches from spectral graph wavelets to graph filter banks, we focus on the two main questions of how to design the spectral filters and how to select the center vertices to which the patterns are localized. Throughout, we emphasize computationally efficient methods that ensure the resulting transforms and their inverses can be applied to data residing on large, sparse graphs. We demonstrate how this class of transform methods can be used in signal processing tasks such as denoising and non-linear approximation, and provide code for readers to experiment with these methods in new application domains.

The conventional signal processing and beam-forming tasks that must be performed in a sonar system are traditionally accomplished by electronic analog and digital means. Conceptually, these signal-processing tasks can be very conveniently accomplished with optical systems, although the required accompanying optical input recording process is usually a practical difficulty. From another point of view, the process of acoustical holography, with optical reconstruction, can be used as a sonar device, i.e., to see objects in water. These two conceptual approaches to sonar have much in common, and are both subject to several fundamental limitations. An important difference between them is the pulse echo-timing ability of conventional sonar, where range information is implied by echo delay, versus the CW nature of a holographic system, where range is obtained by parallactic or focusing effects. Hybrid systems are possible which incorporate the attributes of both approaches.

Sigma delta modulation is a popular technique for high-resolution analog-to-digital conversion and digital-to-analog conversion. We investigate chaotic phenomena in multibit first-order sigma-delta modulators. Particular attention is placed on the occurrence of periodic orbits or limit cycles. These may result in idle tones audible to the listener when sigma-delta modulation is used for audio signal processing. One suggested method of eliminating idle tones is the operation of a sigma delta modulator in the chaotic regime. Unfortunately, chaotic modulation of a first order sigma delta modulator is a poor system for signal processing. We show that minor variations on a traditional first order sigma-delta modulator, together with a multibit implementation, may be used to produce an effective, stable chaotic modulator that accurately encodes the input and helps remove the presence of idle tones. (c) 2001 American Institute of Physics.

The objective of this paper is to identify the dominant coherent structures within cavitating flow around a Clark-Y hydrofoil using two data-driven modal decomposition methods, proper orthogonal decomposition (POD) and dynamic mode decomposition (DMD). A snapshot data sequence is obtained using a large eddy simulation and the interaction between cavitation and the vortex during cloud cavity shedding evolution is investigated. Modal decomposition via POD and DMD indicates that the dominant coherent structures include the large-scale cavity–vortex, re-entrant jet, shear layer, and small-scale vortex in the wake. In addition, the flow field can be reconstructed from the most energetic POD or DMD modes. The errors in the flow reconstructions produced using the first four POD modes, first eight POD modes, and first eight DMD modes are 3.884%, 3.240%, and 3.889%, respectively. Furthermore, transient cavitating flow can be predicted via the DMD method with an error of 8.081%. The largest errors in the reconstructed and predicted results occur mostly in the shear layer, trailing edge, and near wake. POD and DMD provide accurate and practically beneficial techniques for understanding cavitating flow, although substantial challenges remain with regard to predicting this intense nonlinear system.

We propose a scheme for analog-to-digital (A/D) conversion of a signal that is bandlimited most of the time but contains harmonics of a high frequency occasionally. To enable the capture of the high frequency components, such signals are usually oversampled. This framework is of practical interest in the case of electrical signals wherein one encounters high frequency faults or anomalies from time to time, all of which need to be retained for diagnostic purposes. Existing schemes that rely on oversampling for dithered A/D conversion of bandlimited signals are designed for the worst case and do not take into account the fact that the high frequency anomalies occur rarely. We propose an adaptive A/D conversion scheme that identifies the presence or absence of a high frequency component in a given block of samples by formulating it as a hypothesis testing problem. The proposed scheme performs better in terms of compression compared to a scheme designed for the worst case while retaining the same guarantees on the reconstruction error.

Motivated by partial differential equation (PDE) models in signal processing, an l2 approach to derivative calculation is introduced based on sampled data. This approach utilizes regularity constraints on the continuous-domain signal that are already embedded in the PDE model. In particular, the continuous-domain input signal is assumed to belong to a reproducing kernel Hilbert space and the sampling process (ideal or non-ideal) is shown to correspond to an appropriate orthogonal projection. The values of the derivative function are shown to correspond to a set of inner product calculations, giving rise to a minimax solution for an h approximation problem. Several matrix operators are then demonstrated for 1D and 2D cases, found to be superior to the backward-forward difference approach.

Background and ObjectiveMachine learning has become a powerful tool in several computation domains. The most progressive way of machine learning, deep learning, has already surpassed several algorithms designed by human experts. It also applies to the field of biomedical signal processing. However, while many experts produce deep learning models, there is no software platform for signal processing, allowing the convenient use of pre‐trained deep learning models and evaluating them using any inspected signal. This may also hinder understanding, interpretation, and explanation of results. For these reasons, we designed DeepPlayer. It is a plugin for the free signal processing software SignalPlant. The plugin allows loading deep learning models saved in the Open Neural Network Exchange (ONNX) file format and evaluating them on any given signal.MethodsThe DeepPlayer plugin and its graphical user interface were designed in C# programming language and the .NET framework. We used the inference library OnnxRuntime, which supports graphics card acceleration. The inference is executed in asynchronous tasks for a live preview and evaluation of the signals. Model outputs can be exported back to SignalPlant for further processing, such as peak detection or thresholding.ResultsWe developed the DeepPlayer plugin to evaluate deep learning models in SignalPlant. The plugin keeps with SignalPlant's interactive work with signals, such as live preview or easy selection of associated signals. The plugin can load classification or regression models and allows standard pre‐processing and post‐processing methods. We prepared several deep learning models to test the plugin. Additionally, we provide a tutorial training script that outputs an ONNX format model with correctly set metadata information. These, and the source code of the DeepPlayer plugin, are publicly accessible via GitHub and Google Colab service.ConclusionThe DeepPlayer plugin allows running deep learning models easily and interactively. Therefore, experts and non‐AI experts alike can explore and apply deep learning models for (biomedical) signal processing. Its ease of use and interactivity might also contribute to a better understanding and acceptance of AI methods in biomedicine.

The technologies holds the ability to change the world. Current digital era is a product of the evolutionary technologies. It created the necessity to increase the Human Computer Interaction (HCI) and it became one of the most emerging research areas of the decade. HCI is an interface between the users and the system to improve the interaction. HCI concept came into existence in early 1980's. One of the emerging new research area in HCI is Context Aware System (CAS). The technological advancements in HCI created a new outlook in the research of CAS. CAS is a system which understand the user, their surroundings, and location. CAS make this possible by processing the environmental and bio-acoustic. Sound is one of the important media for both humans and animals to communicate and understand information. Bird sound, vehicle sound, wind sound etc. are some of the environmental and bio acoustics. Processing these sounds or signals will help us to create a better performing CAS. This paper profiles a survey study on bird sound classification and identification. Automatic identification of bird sound is one among the difficult task in signal processing. Also, the paper will profile the previous research works on various phases in bird vocalization processing; such as preprocessing, feature selection and classification.

Inference tasks in signal processing are often characterized by the availability of reliable statistical modeling with some missing instance-specific parameters. One conventional approach uses data to estimate these missing parameters and then infers based on the estimated model. Alternatively, data can also be leveraged to directly learn the inference mapping end to end. These approaches for combining partially known statistical models and data in inference are related to the notions of generative and discriminative models used in the machine learning literature <xref ref-type="bibr" rid="ref1" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">[1]</xref> , <xref ref-type="bibr" rid="ref2" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">[2]</xref> , typically considered in the context of classifiers.

Markov random fields are designed to represent structured dependencies among large collections of random variables, and are well-suited to capture the structure of real-world signals. Many fundamental tasks in signal processing (e.g., smoothing, denoising, segmentation etc.) require efficient methods for computing (approximate) marginal probabilities over subsets of nodes in the graph. The marginalization problem, though solvable in linear time for graphs without cycles, is computationally intractable for general graphs with cycles. This intractability motivates the use of approximate ldquomessage-passingrdquo algorithms. This paper studies the convergence and stability properties of the family of reweighted sum-product algorithms, a generalization of the widely used sum-product or belief propagation algorithm, in which messages are adjusted with graph-dependent weights. For pairwise Markov random fields, we derive various conditions that are sufficient to ensure convergence, and also provide bounds on the geometric convergence rates. When specialized to the ordinary sum-product algorithm, these results provide strengthening of previous analyses. We prove that some of our conditions are necessary and sufficient for subclasses of homogeneous models, but not for general models. The experimental simulations on various classes of graphs validate our theoretical results.

Fitting standard geometric elements into measurement data using Least-Squares techniques is a common task in signal processing across various technical applications. However, the application of these well-established but purely data-based methods does not consider potentially available prior knowledge about the measurand of interest or the measuring device. Thus, up to this day, additional information is usually left unused beyond a few academic applications. By applying a Bayesian approach, this prior knowledge can be incorporated into the fitting task, potentially leading to a reduction in overall uncertainty and fragility of the evaluation result. In this study, Bayesian models are proposed for incorporating prior knowledge into circular, linear, and ellipse fitting tasks. The general approaches as well as specific results are compared to the established Total-Least-Squares method within the example of the application of the F-operator in surface texture measurement illustrating the practical benefits of the approach.

Analyzing the magnitude response of a finite-length sequence is a ubiquitous task in signal processing. However, the discrete Fourier transform (DFT) provides only discrete sampling points of the response characteristic. This work introduces bounds on the magnitude response, which can be efficiently computed without additional zero padding. The proposed bounds can be used for more informative visualization and inform whether additional frequency resolution or zero padding is required.

A major task in signal processing is the determination of the significant frequencies of received waveforms. Radar, sonar, and speech processing all employ techniques that rely on a short time frequency representation of a received waveform. This paper uses a backpropagation neural network to extract information about the presence of a target and, if present, the determination of the dominant frequency in the presence of clutter.