Showing papers in "IEEE Signal Processing Magazine in 2022"
TL;DR: In this paper , the authors provide a tutorial on the fundamental properties of the RIS technology from a signal processing perspective, and exemplify how they can be utilized for improved communication, localization, and sensing.
Abstract: Antenna array technology enables directional transmission and reception of wireless signals, for communications, localization, and sensing purposes. The signal processing algorithms that underpin this technology began to be developed several decades ago [1], but it is first with the ongoing deployment of the fifth-generation (5G) wireless mobile networks that it becomes a mainstream technology [2]. The number of antenna elements in the arrays of the 5G base stations and user devices can be measured at the order of 100 and 10, respectively. As the networks shift towards using higher frequency bands, more antennas fit into a given aperture. The 5G developments enhance the transmitter and receiver functionalities, but the wireless channel propagation remains an uncontrollable system. This is illustrated in Fig. 1(a) and its mathematical notation will be introduced later. Transmitted signals with three different frequencies are shown to illustrate the fact that attenuation can vary greatly across frequencies. Looking beyond 5G, the advent of electromagnetic components that can shape how they interact with wireless signals enables partial control of the propagation. A reconfigurable intelligent surface (RIS) is a two-dimensional surface of engineered material whose properties are reconfigurable rather than static [4]. This article provides a tutorial on the fundamental properties of the RIS technology from a signal processing perspective. It is meant as a complement to recent surveys of electromagnetic and hardware aspects [4], [7], [11], acoustics [12], communication theory [13], and localization [8]. We will provide the formulas and derivations that are required to understand and analyze RIS-aided systems using signal processing, and exemplify how they can be utilized for improved communication, localization, and sensing.
71 citations
TL;DR: Self-supervised representation learning as discussed by the authors aims to provide powerful deep feature learning without the requirement of large annotated datasets, thus alleviating the annotation bottleneck that is one of the main barriers to practical deployment of deep learning today.
Abstract: Self-supervised representation learning methods aim to provide powerful deep feature learning without the requirement of large annotated datasets, thus alleviating the annotation bottleneck that is one of the main barriers to practical deployment of deep learning today. These methods have advanced rapidly in recent years, with their efficacy approaching and sometimes surpassing fully supervised pre-training alternatives across a variety of data modalities including image, video, sound, text and graphs. This article introduces this vibrant area including key concepts, the four main families of approach and associated state of the art, and how self-supervised methods are applied to diverse modalities of data. We further discuss practical considerations including workflows, representation transferability, and compute cost. Finally, we survey the major open challenges in the field that provide fertile ground for future work.
35 citations
TL;DR: In this article , a federated machine learning paradigm for training models across multiple edge devices holding local data sets, without explicitly exchanging the data is proposed, which differs from conventional centralized machine learning and poses several core unique challenges and requirements.
Abstract: The dramatic success of deep learning is largely due to the availability of data. Data samples are often acquired on edge devices, such as smartphones, vehicles, and sensors, and in some cases cannot be shared due to privacy considerations. Federated learning is an emerging machine learning paradigm for training models across multiple edge devices holding local data sets, without explicitly exchanging the data. Learning in a federated manner differs from conventional centralized machine learning and poses several core unique challenges and requirements, which are closely related to classical problems studied in the areas of signal processing and communications. Consequently, dedicated schemes derived from these areas are expected to play an important role in the success of federated learning and the transition of deep learning from the domain of centralized servers to mobile edge devices.
32 citations
TL;DR: In this paper , a review of recent advances in incorporating sparsity-promoting priors into three highly popular data modeling tools, namely deep neural networks, Gaussian processes, and tensor decomposition, is presented.
Abstract: Sparse modeling for signal processing and machine learning has been at the focus of scientific research for over two decades. Among others, supervised sparsity-aware learning comprises two major paths paved by: a) discriminative methods and b) generative methods. The latter, more widely known as Bayesian methods, enable uncertainty evaluation w.r.t. the performed predictions. Furthermore, they can better exploit related prior information and naturally introduce robustness into the model, due to their unique capacity to marginalize out uncertainties related to the parameter estimates. Moreover, hyper-parameters associated with the adopted priors can be learnt via the training data. To implement sparsity-aware learning, the crucial point lies in the choice of the function regularizer for discriminative methods and the choice of the prior distribution for Bayesian learning. Over the last decade or so, due to the intense research on deep learning, emphasis has been put on discriminative techniques. However, a come back of Bayesian methods is taking place that sheds new light on the design of deep neural networks, which also establish firm links with Bayesian models and inspire new paths for unsupervised learning, such as Bayesian tensor decomposition. The goal of this article is two-fold. First, to review, in a unified way, some recent advances in incorporating sparsity-promoting priors into three highly popular data modeling tools, namely deep neural networks, Gaussian processes, and tensor decomposition. Second, to review their associated inference techniques from different aspects, including: evidence maximization via optimization and variational inference methods. Challenges such as small data dilemma, automatic model structure search, and natural prediction uncertainty evaluation are also discussed. Typical signal processing and machine learning tasks are demonstrated.
22 citations
TL;DR: In this article , a physics-informed neural network (PINN) is employed to quantify the microstructure of a polycrystalline Nickel by computing the spatial variation of compliance coefficients (compressibility, stiffness and rigidity) of the material.
Abstract: We employ physics-informed neural networks (PINNs) to quantify the microstructure of a polycrystalline Nickel by computing the spatial variation of compliance coefficients (compressibility, stiffness and rigidity) of the material. The PINN is supervised with realistic ultrasonic surface acoustic wavefield data acquired at an ultrasonic frequency of 5 MHz for the polycrystalline material. The ultrasonic wavefield data is represented as a deformation on the top surface of the material with the deformation measured using the method of laser vibrometry. The ultrasonic data is further complemented with wavefield data generated using a finite element based solver. The neural network is physically-informed by the in-plane and out-of-plane elastic wave equations and its convergence is accelerated using adaptive activation functions. The overarching goal of this work is to infer the spatial variation of compliance coefficients of materials using PINNs, which for ultrasound involves the spatially varying speed of the elastic waves. More broadly, the resulting PINN based surrogate model shows a promising approach for solving ill-posed inverse problems, often encountered in the non-destructive evaluation of materials.
22 citations
TL;DR: This article presents a unified and principled review of PnP by tracing its roots, describing its major variations, summarizing main results, and discussing applications in computational imaging.
Abstract: Plug-and-play (PnP) priors constitute one of the most widely used frameworks for solving computational imaging problems through the integration of physical models and learned models. PnP leverages high-fidelity physical sensor models and powerful machine learning methods for prior modeling of data to provide state-of-the-art reconstruction algorithms. PnP algorithms alternate between minimizing a data fidelity term to promote data consistency and imposing a learned regularizer in the form of an image denoiser. Recent highly successful applications of PnP algorithms include biomicroscopy, computerized tomography (CT), magnetic resonance imaging (MRI), and joint ptychotomography. This article presents a unified and principled review of PnP by tracing its roots, describing its major variations, summarizing main results, and discussing applications in computational imaging. We also point the way toward further developments by discussing recent results on equilibrium equations that formulate the problem associated with PnP algorithms.
19 citations
TL;DR: This work has shown that hyperparameters associated with the adopted priors, which correspond to cost function regularizers, can be learned via the training data and not via costly cross-validation techniques, which is, in general, the case with the discriminative methods.
Abstract: Sparse modeling for signal processing and machine learning, in general, has been at the focus of scientific research for over two decades. Among others, supervised sparsity-aware learning (SAL) consists of two major paths paved by 1) discriminative methods that establish direct input–output mapping based on a regularized cost function optimization and 2) generative methods that learn the underlying distributions. The latter, more widely known as Bayesian methods, enable uncertainty evaluation with respect to the performed predictions. Furthermore, they can better exploit related prior information and also, in principle, can naturally introduce robustness into the model, due to their unique capacity to marginalize out uncertainties related to the parameter estimates. Moreover, hyperparameters (tuning parameters) associated with the adopted priors, which correspond to cost function regularizers, can be learned via the training data and not via costly cross-validation techniques, which is, in general, the case with the discriminative methods.
19 citations
TL;DR: Recently, deep learning has become the main research frontier for biological image reconstruction and enhancement problems thanks to their high performance and ultrafast inference times as discussed by the authors . However, due to the difficulty of obtaining matched reference data for supervised learning, there has been increasing interest in unsupervised learning approaches that do not need paired reference data.
Abstract: Recently, deep learning (DL) approaches have become the main research frontier for biological image reconstruction and enhancement problems thanks to their high performance and ultrafast inference times. However, due to the difficulty of obtaining matched reference data for supervised learning, there has been increasing interest in unsupervised learning approaches that do not need paired reference data. In particular, self-supervised learning and generative models have been successfully used for various biological imaging applications. In this article, we provide an overview of these approaches from a coherent perspective in the context of classical inverse problems and discuss their applications to biological imaging, including electron, fluorescence, deconvolution microscopy, optical diffraction tomography (ODT), and functional neuroimaging.
18 citations
TL;DR: A comprehensive survey of DL methods for open-world EEG decoding can be found in this paper , which identifies promising research directions to inspire future studies for EEG decoding in real-world applications.
Abstract: Electroencephalogram (EEG) decoding aims to identify the perceptual, semantic, and cognitive content of neural processing based on non-invasively measured brain activity. Traditional EEG decoding methods have achieved moderate success when applied to data acquired in static, well-controlled lab environments. However, an open-world environment is a more realistic setting, where situations affecting EEG recordings can emerge unexpectedly, significantly weakening the robustness of existing methods. In recent years, deep learning (DL) has emerged as a potential solution for such problems due to its superior capacity in feature extraction. It overcomes the limitations of defining `handcrafted' features or features extracted using shallow architectures, but typically requires large amounts of costly, expertly-labelled data - something not always obtainable. Combining DL with domain-specific knowledge may allow for development of robust approaches to decode brain activity even with small-sample data. Although various DL methods have been proposed to tackle some of the challenges in EEG decoding, a systematic tutorial overview, particularly for open-world applications, is currently lacking. This article therefore provides a comprehensive survey of DL methods for open-world EEG decoding, and identifies promising research directions to inspire future studies for EEG decoding in real-world applications.
15 citations
TL;DR: This paper interprets the duality between R&C as signals and systems, followed by an introduction of their fundamental principles, and elaborate on the two main trends in their technological evolution, namely, the increase of frequencies and bandwidths, and the expansion of antenna arrays.
Abstract: Radar and communications (R&C) as key utilities of electromagnetic (EM) waves have fundamentally shaped human society and triggered the modern information age. Although R&C had been historically progressing separately, in recent decades, they have been converging toward integration, forming integrated sensing and communication (ISAC) systems, giving rise to new highly desirable capabilities in next-generation wireless networks and future radars. To better understand the essence of ISAC, this article provides a systematic overview of the historical development of R&C from a signal processing (SP) perspective. We first interpret the duality between R&C as signals and systems, followed by an introduction of their fundamental principles. We then elaborate on the two main trends in their technological evolution, namely, the increase of frequencies and bandwidths and the expansion of antenna arrays. We then show how the intertwined narratives of R&C evolved into ISAC and discuss the resultant SP framework. Finally, we overview future research directions in this field.
14 citations
TL;DR: Key components of IPADS learning, including signal generation models, basic deep learning network structures, enhanced data generation and learning methods are discussed, making the learning more scalable, explainable, and better protecting privacy.
Abstract: Deep learning (DL) has driven innovation in the field of computational imaging. One of its bottlenecks is unavailable or insufficient training data. This article reviews an emerging paradigm, imaging physics-based data synthesis (IPADS), that can provide huge training data in biomedical magnetic resonance (MR) without or with few real data. Following the physical law of MR, IPADS generates signals from differential equations or analytical solution models, making learning more scalable and explainable and better protecting privacy. Key components of IPADS learning, including signal generation models, basic DL network structures, enhanced data generation, and learning methods, are discussed. Great IPADS potential has been demonstrated by representative applications in fast imaging, ultrafast signal reconstruction, and accurate parameter quantification. Finally, open questions and future work are discussed.
TL;DR: This article discusses DL methods in the context of four diverse and important categories in the neuroimaging field: classification/prediction, dynamic activity/connectivity, multimodal fusion, and interpretation/visualization, and highlights recent progress in each of these categories.
Abstract: Deep learning (DL) has been extremely successful when applied to the analysis of natural images. By contrast, analyzing neuroimaging data presents some unique challenges, including higher dimensionality, smaller sample sizes, multiple heterogeneous modalities, and a limited ground truth. In this article, we discuss DL methods in the context of four diverse and important categories in the neuroimaging field: classification/prediction, dynamic activity/connectivity, multimodal fusion, and interpretation/visualization. We highlight recent progress in each of these categories, discuss the benefits of combining data characteristics and model architectures, and derive guidelines for the use of DL in neuroimaging data. For each category, we also assess promising applications and major challenges to overcome. Finally, we discuss future directions of neuroimaging DL for clinical applications, a topic of great interest, touching on all four categories.
TL;DR: XAIR as discussed by the authors provides theoretical insights and analysis for explainable AI for regression and classification tasks, and provides demonstrations of XAIR on genuine practical regression problems, and finally discuss the challenges remaining for the field.
Abstract: In addition to the impressive predictive power of machine learning (ML) models, more recently, explanation methods have emerged that enable an interpretation of complex non-linear learning models such as deep neural networks. Gaining a better understanding is especially important e.g. for safety-critical ML applications or medical diagnostics etc. While such Explainable AI (XAI) techniques have reached significant popularity for classifiers, so far little attention has been devoted to XAI for regression models (XAIR). In this review, we clarify the fundamental conceptual differences of XAI for regression and classification tasks, establish novel theoretical insights and analysis for XAIR, provide demonstrations of XAIR on genuine practical regression problems, and finally discuss the challenges remaining for the field.
TL;DR: In this article , the authors describe the evolution of beamforming in the last 25 years from convex to nonconvex optimization and optimization to learning approaches, and provide a glimpse into these important signal processing algorithms for a variety of transmit-receive architectures, propagation zones, propagation paths, and multidisciplinary applications.
Abstract: Beamforming is a signal processing technique to steer, shape, and focus an electromagnetic (EM) wave using an array of sensors toward a desired direction. It has been used in many engineering applications, such as radar, sonar, acoustics, astronomy, seismology, medical imaging, and communications. With the advent of multiantenna technologies in, say, radar and communication, there has been a great interest in designing beamformers by exploiting convex or nonconvex optimization methods. Recently, machine learning (ML) is also leveraged for obtaining attractive solutions to more complex beamforming scenarios. This article captures the evolution of beamforming in the last 25 years from convex to nonconvex optimization and optimization to learning approaches. It provides a glimpse into these important signal processing algorithms for a variety of transmit–receive architectures, propagation zones, propagation paths, and multidisciplinary applications.
TL;DR: This review clarifies the fundamental conceptual differences of XAI for regression and classification tasks, establishes novel theoretical insights and analysis for XAIR, provides demonstrations of XAIR on genuine practical regression problems, and discusses challenges remaining for the field.
Abstract: In addition to the impressive predictive power of machine learning (ML) models, more recently, explanation methods have emerged that enable an interpretation of complex nonlinear learning models, such as deep neural networks. Gaining a better understanding is especially important, e.g., for safety-critical ML applications or medical diagnostics and so on. Although such explainable artificial intelligence (XAI) techniques have reached significant popularity for classifiers, thus far, little attention has been devoted to XAI for regression models (XAIR). In this review, we clarify the fundamental conceptual differences of XAI for regression and classification tasks, establish novel theoretical insights and analysis for XAIR, provide demonstrations of XAIR on genuine practical regression problems, and finally, discuss challenges remaining for the field.
TL;DR: In this paper , the authors discuss DL methods in the context of four diverse and important categories in the neuroimaging field: classification/prediction, dynamic activity/connectivity, multimodal fusion, and interpretation/visualization.
Abstract: Deep learning (DL) has been extremely successful when applied to the analysis of natural images. By contrast, analyzing neuroimaging data presents some unique challenges, including higher dimensionality, smaller sample sizes, multiple heterogeneous modalities, and a limited ground truth. In this article, we discuss DL methods in the context of four diverse and important categories in the neuroimaging field: classification/prediction, dynamic activity/connectivity, multimodal fusion, and interpretation/visualization. We highlight recent progress in each of these categories, discuss the benefits of combining data characteristics and model architectures, and derive guidelines for the use of DL in neuroimaging data. For each category, we also assess promising applications and major challenges to overcome. Finally, we discuss future directions of neuroimaging DL for clinical applications, a topic of great interest, touching on all four categories.
TL;DR: In this paper , the authors present gradient-based interpretability methods for explaining the decisions of deep neural networks and evaluate them for their robustness and adversarial robustness in the context of explainability.
Abstract: With the rise of deep neural networks, the challenge of explaining the predictions of these networks has become increasingly recognized. While many methods for explaining the decisions of deep neural networks exist, there is currently no consensus on how to evaluate them. On the other hand, robustness is a popular topic for deep learning research; however, it is hardly talked about in explainability until very recently. In this tutorial paper, we start by presenting gradient-based interpretability methods. These techniques use gradient signals to assign the burden of the decision on the input features. Later, we discuss how gradient-based methods can be evaluated for their robustness and the role that adversarial robustness plays in having meaningful explanations. We also discuss the limitations of gradient-based methods. Finally, we present the best practices and attributes that should be examined before choosing an explainability method. We conclude with the future directions for research in the area at the convergence of robustness and explainability.
TL;DR: In this paper , the authors proposed a change in the detector specifications that may reduce or lose information that could have been advantageous during data analysis during data acquisition, which may result in the loss of useful information.
Abstract: The arrival of direct electron detectors (DEDs) with high frame rates in the field of scanning transmission electron microscopy (TEM) has enabled many experimental techniques that require collection of a full diffraction pattern at each scan position, a field which is subsumed under the name four-dimensional scanning transmission electron microscopy (4D-STEM). DED frame rates approaching 100 kHz require data transmission rates and data storage capabilities that exceed those of the commonly available computing infrastructures. Current commercial DEDs allow the user to make compromises in pixel bit depth, detector binning, or windowing to reduce the per-frame file size and allow higher frame rates. This change in detector specifications requires decisions to be made before data acquisition that may reduce or lose information that could have been advantageous during data analysis.
DOI•
TL;DR: The arrival of direct electron detectors with high frame rates in the field of scanning transmission electron microscopy (TEM) has enabled many experimental techniques that require collection of a full diffraction pattern at each scan position, a field which is subsumed under the name 4D-STEM.
Abstract: The arrival of direct electron detectors (DEDs) with high frame rates in the field of scanning transmission electron microscopy (TEM) has enabled many experimental techniques that require collection of a full diffraction pattern at each scan position, a field which is subsumed under the name four-dimensional scanning transmission electron microscopy (4D-STEM). DED frame rates approaching 100 kHz require data transmission rates and data storage capabilities that exceed those of the commonly available computing infrastructures. Current commercial DEDs allow the user to make compromises in pixel bit depth, detector binning, or windowing to reduce the per-frame file size and allow higher frame rates. This change in detector specifications requires decisions to be made before data acquisition that may reduce or lose information that could have been advantageous during data analysis.
TL;DR: In this paper , the most prominent radio map estimation methods are discussed, starting from simple regression, the exposition gradually delves into more sophisticated algorithms, eventually touching upon state-of-the-art techniques.
Abstract: Radio maps characterize quantities of interest in radio communication environments, such as the received signal strength and channel attenuation, at every point of a geographical region. Radio map estimation (RME) typically entails interpolative inference based on spatially distributed measurements. In this tutorial article, after presenting some representative applications of radio maps, the most prominent RME methods are discussed. Starting from simple regression, the exposition gradually delves into more sophisticated algorithms, eventually touching upon state-of-the-art techniques. To gain insight into this versatile toolkit, illustrative toy examples will also be presented.
TL;DR: An overview of the recent developments in incorporating physics information into learning-based MRI reconstruction, covering physics-driven loss functions, plug-and-play methods, generative models, and unrolled networks is provided.
Abstract: Physics-driven deep learning methods have emerged as a powerful tool for computational magnetic resonance imaging (MRI) problems, pushing reconstruction performance to new limits. This article provides an overview of the recent developments in incorporating physics information into learning-based MRI reconstruction. We consider inverse problems with both linear and nonlinear forward models for computational MRI and review the classical approaches for solving these. We then focus on physics-driven deep learning approaches, covering physics-driven loss functions, plug-and-play (PnP) methods, generative models, and unrolled networks. We highlight domain-specific challenges, such as real- and complex-valued building blocks of neural networks, and translational applications in MRI with linear and nonlinear forward models. Finally, we discuss common issues and open challenges, and we draw connections to the importance of physics-driven learning when combined with other downstream tasks in the medical imaging pipeline.
TL;DR: In this paper , the Cramer-Rao bound (CRB) is used for signal parameter estimation from real-valued quantized data and its derivation was somewhat sketchy.
Abstract: Several current ultra-wide band applications, such as millimeter wave radar and communication systems, require high sampling rates and therefore expensive and energy-hungry analogto-digital converters (ADCs). In applications where cost and power constraints exist, the use of high-precision ADCs is not feasible and the designer must resort to ADCs with coarse quantization. Consequently the interest in the topic of signal parameter estimation from quantized data has increased significantly in recent years. The Cramer-Rao bound (CRB) is an important yardstick in any parameter estimation problem. Indeed it lower bounds the variance of any unbiased parameter estimator. Moreover, the CRB is an achievable limit, for instance it is asymptotically attained by the maximum likelihood estimator (under regularity conditions), and thus it is a useful benchmark to which the accuracy of any parameter estimator can and should be compared. A formula for the CRB for signal parameter estimation from real-valued quantized data has been presented in but its derivation was somewhat sketchy. The said CRB formula has been extended for instance in to complex-valued quantized data, but again its derivation was rather sketchy. The special case of binary (1-bit) ADCs and a signal consisting of one sinusoid has been thoroughly analyzed in . The CRB formula for a binary ADC and a general real-valued signal has been derived.
TL;DR: In applications where cost and power constraints exist, the use of high-precision ADCs is not feasible, and the designer must resort to ADCs with coarse quantization.
Abstract: Several current ultrawide band applications, such as millimeter-wave radar and communication systems [1] –[3] , require high sampling rates and therefore expensive and energy-hungry analog-to-digital converters (ADCs). In applications where cost and power constraints exist, the use of high-precision ADCs is not feasible, and the designer must resort to ADCs with coarse quantization. Consequently, the interest in the topic of signal parameter estimation from quantized data has increased significantly in recent years.
TL;DR: In this paper , the authors systematically review the integration of fundamental and advanced signal processing techniques for AR/MR audio to equip researchers and engineers in the signal processing community for the next wave of augmented or mixed reality applications.
Abstract: Augmented or mixed reality (AR/MR) is emerging as one of the key technologies in the future of computing. Audio cues are critical for maintaining a high degree of realism, social connection, and spatial awareness for various AR/MR applications, such as education and training, gaming, remote work, and virtual social gatherings to transport the user to an alternate world called the metaverse. Motivated by a wide variety of AR/MR listening experiences delivered over hearables, this article systematically reviews the integration of fundamental and advanced signal processing techniques for AR/MR audio to equip researchers and engineers in the signal processing community for the next wave of AR/MR.
TL;DR: In this article , the authors present an overview of signal processing techniques for joint terahertz communications and sensing (CAS) applications, with an emphasis on signal preprocessing and feature extraction.
Abstract: Following the recent progress in terahertz (THz) signal generation and radiation methods, joint THz communications and sensing (CAS) applications are being proposed for future wireless systems. Toward this end, THz spectroscopy is expected to be carried over user equipment devices to identify material and gaseous components of interest. THz-specific signal processing techniques should complement this resurgent interest in THz sensing for efficient utilization of the THz band. In this article, we present an overview of these techniques, with an emphasis on signal preprocessing [standard normal variate (SNV) normalization, minimum–maximum normalization, and Savitzky–Golay (SG) filtering], feature extraction [principal component analysis (PCA), partial least squares (PLS), t -distributed stochastic neighbor embedding ( t -SNE), and nonnegative matrix factorization (NMF)], and classification techniques [support vector machines (SVMs), the k -nearest neighbor ( k NN), discriminant analysis (DA), and naive Bayes (NB)]. We also address the effectiveness of deep learning techniques by exploring their promising sensing and localization capabilities at the THz band. Finally, we investigate the performance and complexity tradeoffs of the studied methods in the context of joint CAS (JCAS). We thereby motivate corresponding use cases and present a handful of contextual future research directions.
TL;DR: This article systematically reviews the integration of fundamental and advanced signal processing techniques for AR/MR audio to equip researchers and engineers in the signal processing community for the next wave of AR/ MR.
Abstract: Augmented or mixed reality (AR/MR) is emerging as one of the key technologies in the future of computing. Audio cues are critical for maintaining a high degree of realism, social connection, and spatial awareness for various AR/MR applications, such as education and training, gaming, remote work, and virtual social gatherings to transport the user to an alternate world called the metaverse. Motivated by a wide variety of AR/MR listening experiences delivered over hearables, this article systematically reviews the integration of fundamental and advanced signal processing techniques for AR/MR audio to equip researchers and engineers in the signal processing community for the next wave of AR/MR.
TL;DR: The method of deep unrolling is reviewed and it is shown how it improves source localization in several biological imaging settings.
Abstract: Deep algorithm unrolling has emerged as a powerful, model-based approach to developing deep architectures that combine the interpretability of iterative algorithms with the performance gains of supervised deep learning, especially in cases of sparse optimization. This framework is well suited to applications in biological imaging, where physics-based models exist to describe the measurement process and the information to be recovered is often highly structured. Here we review the method of deep unrolling and show how it improves source localization in several biological imaging settings.
TL;DR: This work states that interpretability, reproducibility, and, ultimately, the ability to generalize these solutions to unseen scenarios and situations are all strongly tied to the starting point of explainability.
Abstract: Data-driven solutions are playing an increasingly important role in numerous practical problems across multiple disciplines. The shift from the traditional model-driven approaches to those that are data driven naturally emphasizes the importance of the explainability of solutions, as, in this case, the connection to a physical model is often not obvious. Explainability is a broad umbrella and includes interpretability, but it also implies that the solutions need to be complete, in that one should be able to “audit” them, ask appropriate questions, and hence gain further insight about their inner workings [1]. Thus, interpretability, reproducibility, and, ultimately, our ability to generalize these solutions to unseen scenarios and situations are all strongly tied to the starting point of explainability.
TL;DR: In this article , the authors focus on the importance of explainability of data-driven solutions, as, in this case, the connection to a physical model is often not obvious.
Abstract: Data-driven solutions are playing an increasingly important role in numerous practical problems across multiple disciplines. The shift from the traditional model-driven approaches to those that are data driven naturally emphasizes the importance of the explainability of solutions, as, in this case, the connection to a physical model is often not obvious. Explainability is a broad umbrella and includes interpretability, but it also implies that the solutions need to be complete, in that one should be able to “audit” them, ask appropriate questions, and hence gain further insight about their inner workings [1] . Thus, interpretability, reproducibility, and, ultimately, our ability to generalize these solutions to unseen scenarios and situations are all strongly tied to the starting point of explainability.
TL;DR: Deep algorithm unrolling has emerged as a powerful, model-based approach to developing deep architectures that combine the interpretability of iterative algorithms with the performance gains of supervised deep learning, especially in cases of sparse optimization as mentioned in this paper .
Abstract: Deep algorithm unrolling has emerged as a powerful, model-based approach to developing deep architectures that combine the interpretability of iterative algorithms with the performance gains of supervised deep learning, especially in cases of sparse optimization. This framework is well suited to applications in biological imaging, where physics-based models exist to describe the measurement process and the information to be recovered is often highly structured. Here we review the method of deep unrolling and show how it improves source localization in several biological imaging settings.