Acoustic vector sensor (AVS) has been recently researched and developed for acoustic wave capturing signal processing. Conventional array generally employs spatially displayed sensors enhancement, source localisation, target tracking, etc. However, the large size usually limits its implementations on some portable devices. AVS which includes one omni-directional three orthogonally co-located directional introduced. An is able to provide four-dimensional information of sound field in space:...

Seizure onset detection and epileptic preictal prediction based on electroencephalogram (EEG) signals have been a challenge problem in the research community. In this brief, novel states classification algorithm multichannel EEGs representation using multiple hand-crafted features, feature fusion transfer learning (TL) with pre-trained deep neural networks (DNNs), discriminative extraction state hierarchical network (HNN), is developed. The mean amplitude spectrum (MAS), power spectral...

Constrained least-squares design and constrained Chebyshev of one- two-dimensional nonlinear-phase FIR filters with prescribed phase error are considered in this paper by a unified semi-infinite positive-definite quadratic programming approach. In order to obtain unique optimal solutions, we propose impose constraints on the complex approximation error. By introducing sigmoid phase-error constraint bound function, group-delay can be greatly reduced. A Goldfarb-Idnani based algorithm is...

The extreme learning machine (ELM) has attracted much attention over the past decade due to its fast speed and convincing generalization performance. However, there still remains a practical issue be approached when applying ELM: randomly generated hidden node parameters without tuning can lead outputs being nonuniformly distributed, thus giving rise poor To address this deficiency, novel activation function with an affine transformation (AT) on input is introduced into ELM, which leads...

The scalp electroencephalogram (EEG) has been extensively studied for epileptic signal classification in the past, but little attention paid to data imbalance among different states. It is well known that, general, duration of seizure onset less than several minutes or even shorter. This will result an problem when comparing durations preictal and interictal In this article, a novel detection algorithm imbalanced proposed. wavelet packet decomposition (WPD)-based statistical features (SFs)...

The minimax design of infinite impulse response (IIR) digital filters is a nonconvex optimization problem, and thus has many local minima. It shown in this correspondence that sequential constrained least-squares (SCLS) method higher possibility obtaining better solutions than direct minimization when applied to the IIR filters. We combine SCLS with Steiglitz-McBride (SM) strategy, resulting practical procedure. positive realness stability condition proposed by Dumitrescu Niemisto...

Linear phase is an important characteristic of digital filters in many signal processing applications. In this paper, two iterative reweighted minimax error algorithms are proposed to design nearly linear-phase infinite impulse response (IIR) with prescribed or simultaneously minimized magnitude errors and preset transition-band gain. each iteration the algorithms, a weighted problem fixed weight function firstly solved using modified Gauss-Newton method variable step length, then updated...

One of the salient features extreme learning machine (ELM) is its fast speed. However, in a big data environment, ELM still suffers from an overly heavy computational load due to high dimensionality and large amount data. Using alternating direction method multipliers (ADMM), convex model fitting problem can be split into set concurrently executable subproblems, each with just subset coefficients. By maximally splitting across coefficients incorporating novel relaxation technique, relaxed...

Autoencoding is a vital branch of representation learning in deep neural networks (DNNs). The extreme machine-based autoencoder (ELM-AE) has been recently developed and gained popularity for its fast speed ease implementation. However, the ELM-AE uses random hidden node parameters without tuning, which may generate meaningless encoded features. In this brief, we first propose within-class scatter information constraint-based AE (WSI-AE) that minimizes both reconstruction error We then build...

In this paper, a voting based weighted online sequential extreme learning machine (VWOS-ELM) is proposed for class imbalance (CIL). VWOS-ELM the first classifier that can tackle problem in multi-class data streams. Utilizing WOS-ELM and recently (VOS-ELM) method, adapts better to newly received than original method. Experimental results show outperforms both recent meta-cognitive methods. It also achieves similar performance of ensemble subset OS-ELM (ESOS-ELM) but using fewer independent...

Angles of arrival (AOAs), gain ratios (GROAs), and time differences (TDOAs) are the three most commonly used signal metrics for source localization in acoustic sensor array networks. It is intuitive to expect a performance increase by combining those metrics. In this paper, we develop feasible new positioning framework using joint AOA-GROA-TDOA measurements, establish Cramér-Rao lower bound (CRLB) method quantify compared with existing methods. Our analysis starts from received waveforms...

The epileptic state classification for preictal prediction of seizure is essential preventing the damages caused by seizure. Existing research mainly focus on detection while few them investigate prediction. In this paper. a novel electroencephalograph signal (EEG) based algorithm developed. one hour before divided into three non-overlapped segments and multi-class then performed wavelet packet features (WPFs), including subband energy ratio entropies, are extracted multi-channel EEG...

Infinite impulse response (IIR) digital filters with prescribed magnitude and phase responses have been used in many applications. To approximate the responses, we propose a new approach to design of general IIR by minimizing maximum error subject or simultaneously minimized error, where are controlled two elliptic constraints respectively major minor axes along desired frequency response. The sequential constrained least-squares method Levy-Sanathanan-Koerner strategy convert nonconvex into...

Finite impulse response (FIR) filters without phase specifications have found many applications in signal processing and communications. It is shown this paper that the constrained <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">${L}_{{p}}$</tex></formula> magnitude error design of such an FIR filter equivalent to a xmlns:xlink="http://www.w3.org/1999/xlink"> <tex approximation given with equal...