In this letter, we propose a vocal tract length (VTL) perturbation method for text-dependent speaker verification (TD-SV), in which a set of TD-SV systems are trained, one for each VTL factor, and score-level fusion is applied to make a final decision. Next, we explore the bottleneck (BN) feature extracted by training deep neural networks with a self-supervised learning objective, autoregressive predictive

This work generalizes the Fourier transform (FT)-based frequency-invariant beamforming (FIB) method to the synthesis of scanned frequency-invariant (FI) beam pattern with space-frequency notching for an array with non-isotropic elements. Wideband FI pattern characteristics are described by using multiple reference sub-band FI patterns. By applying fast Fourier transform (FFT) on the combination of

In this work, we study the impact of input-spread on the steady-state excess mean squared error (EMSE) of the normalized least mean squares (NLMS) algorithm. First, we use the concept of majorization to order the input-regressors according to their spread. Second, we use Schur-convexity to show that the majorization order of the input-regressors is preserved in the EMSE. Effectively, we provide an

We present a framework for the optimal filtering of spherical signals contaminated by realizations of an additive, zero-mean, uncorrelated and anisotropic noise process on the sphere. Filtering is performed in the wavelet domain given by the scale-discretized wavelet transform on the sphere. The proposed filter is optimal in the sense that it minimizes the mean square error between the filtered wavelet

We present a new algorithm for estimating the Point Spread Function (PSF) in wide-field astronomical images with extreme source crowding. Robust and accurate PSF estimation in crowded astronomical images dramatically improves the fidelity of astrometric and photometric measurements extracted from wide-field sky monitoring imagery. Our radically new approach utilizes convolutional sparse representations

The main focus of unsupervised person re-identification is the clustering of unlabeled samples in the target domain. However, most existing studies neglected to mine the deep semantic information of the target domain and did not consider a better combination of the source domain and the target domain. In this letter, we not only consider the changes of the target domain within its own domain but also

Joint power optimization of sources and relay in cooperative networks, while multiple source-destination (SD) pairs are utilizing a relay and also have direct transmission links, is addressed in this letter. In the assumed network each SD pair requires a minimum quality of service (QoS) in terms of its received SINR. We have shown that the resulted power control equations are quadratic due to coexistence

Realistic low-resolution (LR) face images refer to those captured by the real-world surveillance cameras at extreme standoff distances, thereby LR and poor in quality essentially. Owing to severe scarcity of labeled data, a high-capacity deep convolution neural networks (CNN) is hardly trained to confront the realistic LR face recognition (LRFR) challenge. We introduce in this letter a dual-stream

Mutually orthogonal complementary sets (MOCSs) have found a number of practical applications in wireless communications and radar owing to their perfect aperiodic auto-correlation and cross-correlation properties. Recently, toward the challenge of designing MOCSs with non-power-of-two lengths, direct constructions were presented by Wu et al. using generalized Boolean functions (GBFs). In this letter

Nonnegative matrix factorization (NMF) has been traditionally considered a promising approach for audio source separation. While standard NMF is only suited for single-channel mixtures, extensions to consider multi-channel data have been also proposed. Among the most popular alternatives, multichannel NMF (MNMF) and further derivations based on constrained spatial covariance models have been successfully

Image distortion is a main challenge for tasks on panoramas. In this work, we propose a Distortion-Aware Monocular Omnidirectional (DAMO) network to estimate dense depth maps from indoor panoramas. First, we introduce a distortion-aware module to extract semantic features from omnidirectional images. Specifically, we exploit deformable convolution to adjust its sampling grids to geometric distortions

Filter pruning for a pre-trained convolutional neural network is most normally performed through human-made constraints or criteria such as norms, ranks, etc. Typically, the pruning pipeline comprises two-stage: first learn a sparse structure from the original model, then optimize the weights in the new prune model. One disadvantage of using human-made criteria to prune filters is that the design and

In this letter, a robust minimum error entropy based cubature information filter is proposed for state estimation in non-Gaussian measurement noise. A new combined optimization cost is defined based on the error entropy. Through cubature transform, a statistical linearization regression model is constructed, and a new information filter is then developed by minimizing the error entropy based cost.

This letter proposes a new travel time estimation model based on graph neural network (GraphTTE) to improve the accuracy of travel time estimation. We design a Multi-layer Spatiotemporal Graph frame (MSG), which consists of static network and dynamic networks, to fully consider the influence of traffic temporal characteristics and road network topological characteristics on travel time. Moreover, we

Fine-grained sketch-based image retrieval (FG-SBIR) is an emerging topic in high-level computer vision. Among existing methods, edge-only information based ones are convenient to use but incompetent in distinguishing among ambiguous samples. On the other hand, even though the methods using additional color information can significantly improve the retrieval performance, they are not convenient for

The classical multi-armed bandit (MAB) framework studies the exploration-exploitation dilemma of the decision-making problem and always treats the arm with the highest expected reward as the optimal choice. However, in some applications, an arm with a high expected reward can be risky to play if the variance is high. Hence, the variation of the reward should be considered to make the arm-selection

This letter considers the design of a two-stage direction of arrival (DOA) scheme for radar systems. Precisely, at the first stage a sparse recovery approach is used to obtain both DOA and complex amplitude estimates of the incoming signal. Since the DOA is evaluated on a predefined grid of bins sampling the antenna azimuth mainbeam, at the second stage, a closed-form complex-valued parabolic interpolation

Siamese tracking methods have become the focus of visual tracking in recent years. Advanced Siamese trackers perform well on certain benchmarks, but there are still some limitations. First, most Siamese trackers adopt the initial frame as a single template, which leads to underfitting and reduces the ability to predict instances. Second, mainstream trackers report a rectangular bounding box as a prediction

The randomized subspace Newton convex methods for the sensor selection problem are proposed. The randomized subspace Newton algorithm is straightforwardly applied to the convex formulation, and the customized method in which the part of the update variables are selected to be the present best sensor candidates is also considered. In the converged solution, almost the same results are obtained by original

A short train of equidistant samples rarely suffices to compute a reasonable approximation to a continuous-time signal. Hence, there arises a need for signal reconstruction methods that exploit multiple trains of samples. If the starting times of the trains are known, then stroboscopic techniques can be used for signal reconstruction. If these times are not given, i.e., in the case of asynchronous

A binary ensemble classification method that sequentially processes the data collected from multiple decision agents in the presence of parameter uncertainties is proposed. Agents are assumed to form correlated groups whose decisions are modeled as multivariate Bernoulli random vectors. The prior probabilities of the binary hypotheses and the corresponding probabilities of the outcomes under each hypothesis

Low power requirements, real-time constraints and big data applications present several challenges to digital signal processing (DSP) that span multiple domains including filtering and frequency analysis. Design of digital filters that can efficiently process signals with low computational complexity is desirable and requires innovative approaches. In this paper, we present an efficient two-step design

A novel spherical convolution is defined through the sifting property of the Dirac delta on the sphere. The so-called sifting convolution is defined by the inner product of one function with a translated version of another, but with the adoption of an alternative translation operator on the sphere. This translation operator follows by analogy with the Euclidean translation when viewed in harmonic space

A popular approach to estimate a source location using time difference of arrival (TDOA) measurements is to construct an objective function based on the maximum likelihood (ML) method. An iterative algorithm can be employed to minimize that objective function. The main challenge in this optimization process is the non-convexity of the objective function, which precludes the use of many standard convex

This letter studies a new expectation maximization (EM) algorithm to solve the problem of circle, sphere and more generally hypersphere fitting. This algorithm relies on the introduction of random latent vectors having a priori independent von Mises-Fisher distributions defined on the hypersphere. This statistical model leads to a complete data likelihood whose expected value, conditioned on the observed

A multiple target tracking filter is developed for merged measurement problems that arise with finite resolution sensors. The resulting combinatorial problem is incorporated directly in the joint likelihood function using analytic combinatorics techniques. The Bayesian filter is a sum of several terms that correspond one-to-one to the set of all feasible hypotheses about measurements and targets, i

This letter focuses on the compression of deep features. With the rapid expansion of deep feature data in various CNN-based analysis and processing tasks, the demand for efficient compression continues to increase. Product quantization (PQ) is widely used in the compact expression of features. In the quantization process, feature vectors are mapped into fixed-length codes based on a pre-trained codebook

In this work, we present an energy-efficient distributed learning framework using low-resolution ADCs and coarsely quantized signals for Internet of Things (IoT) networks. In particular, we develop a distributed quantization-aware least-mean square (DQA-LMS) algorithm that can learn parameters in an energy-efficient fashion using signals quantized with few bits while requiring a low computational cost

Benefited from the powerful ability of spatial temporal Graph Convolutional Networks (ST-GCNs), skeleton-based human action recognition has gained promising success. However, the node interaction through message propagation does not always provide complementary information. Instead, it May even produce destructive noise and thus make learned representations indistinguishable. Inevitably, the graph

Average consensus algorithms have wide applications in distributed computing systems where all the nodes agree on the average value of their initial states by only exchanging information with their local neighbors. In this letter, we look into link-based network metrics which are polynomial functions of pair-wise node attributes defined over the links in a network. Different from node-based average

In this work, we leverage on a novel distributional loss to improve vector-to-vector regression for feature-based speech enhancement (SE). The distributional loss function is devised based on the Kullback-Leibler divergence between a selected target distribution and a conditional distribution to be learned from the data for each coefficient in the clean speech vector given the noisy input features

In this paper, the problem of blind equalization of high-order quadrature amplitude modulation (QAM) signals is tackled by using a batch equalizer based on support vector regression (SVR). A new set of error functions weighted by neighborhood symbol decisions and augmented by generalized power factors $p$ and $q$ , are proposed to be used as the penalty terms in SVR, and the optimal values of $p$ and

Deep neural networks have achieved great progress in 3D scene understanding. However, recent methods mainly focused on objects with canonical orientations in contrast with random postures in reality. In this letter, we propose a hierarchical neural network, named Local Frame Network (LFNet), based on the local rotation invariant coordinate frame for robust point cloud analysis. The local point patches

The performance of sparse signal recovery (SSR) can be enhanced by exploiting rich temporal correlation in the multiple snapshots of signal of interest. However, existing methods need to transform the temporally correlated multiple measurements SSR problem into its vectorization form, imposing huge computational cost for algorithmic realization. To overcome this drawback, we propose a novel formulation

Classification of electroencephalogram-based motor imagery (MI-EEG) tasks is crucial in brain computer interfaces (BCI). In view of the characteristics of non-stationarity, time-variability and individual diversity of EEG signals, a novel framework based on graph neural network is proposed for MI-EEG classification. First, an adaptive graph convolutional layer (AGCL) is constructed, by which the electrode

In the literature, the task of dysarthric speech intelligibility assessment has been approached through development of different low-level feature representations, subspace modeling, phone confidence estimation or measurement of automatic speech recognition system accuracy. This paper proposes a novel approach where the intelligibility is estimated as the percentage of correct words uttered by a speaker

A central problem in Binary Hypothesis Testing (BHT) is to determine the optimal tradeoff between the Type I error (referred to as false alarm ) and Type II (referred to as miss ) error. In this context, the exponential rate of convergence of the optimal miss error probability — as the sample size tends to infinity — given some (positive) restrictions on the false alarm probabilities is a fundamental

Although deep neural networks have made significant progress in semantic segmentation, speed and computational cost still can’t meet the strict requirements of real-world applications. In this paper, we present an enhanced asymmetric convolution network (EACNet) to seek a balance between accuracy and speed. Specifically, we design a pair of enhancing asymmetric convolution modules constructed by depth-wise

This letter proposes a robust detector based on kernel correlation (KC) for detecting the presence of a common random signal shared in two channels corrupted by correlated non-Gaussian impulsive noise. A bivariate Gaussian mixture (GM) distribution is employed to simulate the correlation and impulsive characteristic of the noise across two channels. The test statistic is constructed by the dot product

The recent end-to-end co-saliency models have good performance, however, they cannot express the semantic consistency among a group of images well and usually require many co-saliency labels. To this end, a two-stream encoder generative adversarial network (TSE-GAN) with progressive training is proposed in this paper. In the pre-training stage, the salient object detection generative adversarial networks

Recently, deep learning has become a mainstream method of image inpainting. It can not only restore the image texture, obtain high-level abstract features of images, but also restore semantic images such as human face images. Among these methods, generative adversarial networks (GANs) using autoencoder as the generator have become the promising model for image inpainting. These models implement the

Approximate message passing (AMP) is an iterative signal recovery algorithm for compressed sensing (CS) applications. In this letter, we present an integral-based orthogonal AMP (IB-OAMP) technique that avoids the requirements of AMP (and also the original form of OAMP) on differentiable and separable denoisers. The orthogonality in IB-OAMP can be established using a Monte Carlo method similar to the

Underwater images suffer from severe color casts, low contrast and blurriness, which are caused by scattering and absorption when light propagates through water. However, existing deep learning methods treat the restoration process as a whole and do not fully consider the underwater physical distortion process. Thus, they cannot adequately tackle both absorption and scattering, leading to poor restoration

In this letter, a method called Semantic Information Oriented No-Reference (SIONR) video quality assessment model is developed, which can effectively represent quality degradation of video by taking the variations of semantic information into consideration. Specially, temporal variations of the semantic features between adjacent frames are calculated to consider the inconsistency of the static semantic

Contrast enhancement (CE) of magnetic resonance (MR) brain images is an important technique to bring out the tissue details for clinical diagnosis. Recently, a new form of image enhancement has been proposed to complete the task without any information loss. Specifically, information required to restore the original image is reversibly hidden into the enhanced image. Moreover, several image segmentation

We propose a new procedure for image illumination correction with data-driven parameter choice. This procedure aims at estimating the reflectance image from a corrupted version in which the corruption is due to pointwise multiplicative illumination artifact. The $\log \text{-illumination}$ artefact consists of “smooth” variations of the intensity which are modelled by a function lying in a finite dimensional

We consider the problem of robust no reference (NR) video quality assessment (VQA) where the algorithms need to have good generalization performance when they are trained and tested on different datasets. We specifically address this question in the context of predicting video quality for compression and transmission applications. Motivated by the success of the spatio-temporal entropic differences

Estimation after model selection refers to the problem where the exact observation model is unknown and is assumed to belong to a set of candidate models. Thus, a data-based model-selection stage is performed prior to the parameter estimation stage, which affects the performance of the subsequent estimation. In this letter, we investigate post-model-selection Bayesian parameter estimation of a random

The Fractional Fourier Transform (FRFT) has been playing a unique and increasingly important role in signal and image processing. In this letter, we investigate the property of frequency shift in two-dimensional FRFT (2D-FRFT) domain. It is shown that the magnitude of image reconstruction from phase information is frequency shift-invariant in 2D-FRFT domain, enhancing the robustness of image encryption

Component analysis (CA) is a powerful technique for learning discriminative representations in various computer vision tasks. Typical CA methods are essentially based on the covariance matrix of training data. But, the covariance matrix has obvious disadvantages such as failing to model complex relationship among features and singularity in small sample size cases. In this letter, we propose a general

A novel approach is proposed to $H_2$ finite impulse response (FIR) prediction in discrete-time state-space. The biased-constrained $H_2$ optimal unbiased FIR ( $H_2$ -OUFIR) predictor derived under disturbances and measurement errors is shown to have the maximum likelihood form and be equivalent to the OUFIR predictor under Gaussian noise. The derivation is provided using the backward Euler method

Atomic norm minimization is a convex optimization framework to recover point sources from a subset of their low-pass observations, or equivalently the underlying frequencies of a spectrally-sparse signal. When the amplitudes of the sources are positive, a positive atomic norm can be formulated, and exact recovery can be ensured without imposing a separation between the sources, as long as the number

We study the statistical and computational properties of a network Lasso method for local graph clustering. The clusters delivered by nLasso can be characterized elegantly via network flows between cluster boundaries and seed nodes. While spectral clustering methods are guided by a minimization of the graph Laplacian quadratic form, nLasso minimizes the total variation of cluster indicator signals

Robust principal component analysis (RPCA) is a widely used tool for dimension reduction. In this work, we propose a novel non-convex algorithm, coined Iterated Robust CUR (IRCUR), for solving RPCA problems, which dramatically improves the computational efficiency in comparison with the existing algorithms. IRCUR achieves this acceleration by employing CUR decomposition when updating the low rank component

Very deep transformers outperform conventional bi-directional long short-term memory networks for automatic speech recognition (ASR) by a significant margin. However, being autoregressive models, their computational complexity is still a prohibitive factor in their deployment into production systems. To amend this problem, we study two different non-autoregressive transformer structures for ASR: Audio-Conditional

We propose a novel NMF algorithm, named Total Variation constrained Graph-regularized Convex Non-negative Matrix Factorization (TV-GCNMF), to incorporate total variation and graph Laplacian with convex NMF. In this model, the feature details of the data are preserved by a diffusion coefficient based on the gradient information. The graph regularization and convex constraints reveal the intrinsic geometry

Compressed sensing deals with recovery of sparse signals from low dimensional projections, but under the assumption that the measurement setup has infinite dynamic range. In this letter, we consider a system with finite dynamic range, and to counter the clipping effect, the measurements crossing the range are folded back into the dynamic range of the system through modulo arithmetic. For this setup

Sub-graph entropy has recently been applied to functional brain network analysis for identifying important brain regions associated with different brain states and for discriminating brain networks of subjects with psychiatric disorders from healthy controls. This letter describes two pertinent properties of sub-graph entropy. It is shown that when a graph is divided into multiple smaller graphs, the

Differential microphone arrays (DMAs), which are responsive to the differential acoustic pressure fields, have been used in a wide range of applications related to audio and speech. The core part of a DMA is the so-called differential beamformer, which is generally designed by placing a number of nulls in its beampattern to attenuate noise from some directions. But the presence of these nulls may cause

Consider a robust multiple-input single-output downlink beamforming optimization problem in a frequency division duplexing system. The base station (BS) sends training signals to the users, and every user estimates the channel coefficients, quantizes the gain and the direction of the estimated channel and sends them back to the BS. Suppose that the channel state information at the transmitter is imperfectly