We present a system that detects human falls in the home environment, distinguishing them from competing noise, by using only audio signal single far-field microphone. The proposed models each fall or noise segment means of Gaussian mixture model (GMM) supervector, whose Euclidean distance measures pairwise difference between segments. A support vector machine built on kernel GMM supervectors is employed to classify segments into and various types noise. Experiments dataset falls, collected...

This article presents a novel broadband dual-polarized filtering patch antenna array using low-temperature cofired ceramic (LTCC) process for 5G millimeter-wave (mmW) applications. The element is implemented ± 45° polarized fed by two pairs of differential L-type probes. An additional square ring and open strips are connected to each probe, resulting in radiation nulls realizing the lower/upper stopband rejection level over 24 dB with skirt selectivity. cross coupling among these structures...

Along with the advances in multimedia and Internet technology, a huge amount of data, including digital video audio, are generated daily. Tools for efficient indexing retrieval such data indispensable. With multi-modal information present effective integration is necessary still challenging problem. In this paper, we four different methods integrating audio visual classification based on hidden Markov model (HMM): direct concatenation, product HMM, two-stage by neural network. Our results...

Personal recognition using palm–vein patterns has emerged as a promising alternative for human because of its uniqueness, stability, live body identification, flexibility, and difficulty to cheat. With the expanding application pattern recognition, corresponding growth database resulted in long response time. To shorten time this paper proposes simple useful classification identification based on principal direction features. In registration process, Gaussian-Radon transform is adopted...

An event sequence generated by a temporal point process is often associated with hidden and structured branching that captures the triggering relations between its historical current events. In this study, we design new plug-and-play module based on Bregman ADMM (BADMM) algorithm, which infers branches sequences in maximum likelihood estimation framework of processes (TPPs). Specifically, formulate inference as an optimization problem for transition matrix under sparse low-rank constraints,...

Loudspeakers and headsets are ubiquitous electronic devices essential to modern life, finding extensive applications in medicine, education, entertainment, communication. This paper explores their fundamental operating principles, rooted electromagnetic acoustic theories, provides a comprehensive analysis of mechanisms. It summarizes various optimization strategies aimed at enhancing performance, including advancements output, material innovations like carbon nanotubes, structural design...

An event sequence generated by a temporal point process is often associated with hidden and structured branching that captures the triggering relations between its historical current events. In this study, we design new plug-and-play module based on Bregman ADMM (BADMM) algorithm, which infers branches sequences in maximum likelihood estimation framework of processes (TPPs). Specifically, formulate inference as an optimization problem transition matrix under sparse low-rank constraints,...

Estimating road traffic noise is essential for examining the quality of sounding environment and mitigating such a non-negligible pollutant in urban areas. However, existing estimated models often have limited applicability to specific conditions, while required parameters may not be readily available city-wide collection. This paper proposes data-driven approach measuring road-level acoustic information with street view imagery. Specifically, we utilize portable vehicle-equipped hardware...

Background Acute abdominal pain (AAP) is a common symptom presented in the emergency department (ED), and it crucial to have objective accurate triage. This study aims develop machine learning-based prediction model for AAP The goal identify triage indicators critically ill patients ensure prompt availability of diagnostic treatment resources. Methods In this study, we conducted retrospective analysis medical records admitted ED Wuhan Puren Hospital with acute 2019. To high-risk factors,...

Among the existing registration methods, most feature descriptors are designed with image intensity, gradient information and phase congruency (PC). However, both intensity sensitive to illumination changes, complex differences, noise, etc. Despite fact that PC is invariant contrast, it does not perform well when images corrupted noise nonlinear radiation distortions. In this article, we propose a novel called spectrum (SC), which robust variations of intensity. SC focuses on exploiting...

The joint use of hyperspectral image (HSI) and Light Detection And Ranging (LiDAR) data has been widely applied for land cover classification because it can comprehensively represent the urban structures material properties. However, existing methods fail to combine different information effectively, which limits semantic relevance sources. To solve this problem, in paper, an Attention-guided Fusion Classification framework based on Convolutional Neural Network (AFC-CNN) is proposed classify...

Color and shape features have difficulty to recognize fire from fire-colored, irregular-shaped moving objects. An effective algorithm using texture analysis with the Local Binary Pattern (LBP) is proposed in our paper help deal this problem. Besides grayscale rotation invariance, ability of LBP operator for multi-resolution enhanced through multiple operators varying parameters pyramid based image sub-sampling. Experiments are performed on sample pictures red fire, car banner taking...

Abstract Background . Computed tomography perfusion (CTP) imaging plays a critical role in the acute stroke syndrome assessment due to its widespread availability, speed of image acquisition, and relatively low cost. However, repeated scanning protocol, CTP involves substantial radiation dose, which might increase potential cancer risks. Methods In this work, we present novel deep learning model called non-local texture network (NPTN) for high-quality at low-dose cases. Specifically,...

Abstract With more emphasis placed on the transferable acquisition and sustainable development of knowledge data, comprehensiveness, accuracy, scientificity databases are further called for. However, how to evaluate effectiveness database needs research. Thus, this paper builds an integrated evaluation model adopting latent Dirichlet allocation entropy weight method. First, we calculate contents index system by model, which integrates core orientation with characteristics target database....

The low-resolution ultrasound images have poor visual effects. Herein we propose a method for generating clearer intravascular based on super-resolution reconstruction combined with generative adversarial networks. We used the networks to generate by generator and estimate authenticity of discriminator. Specifically, image was passed through sub-pixel convolution layer r2-feature channels maps in same size, followed realignment corresponding pixels each feature map into r ×r sub-blocks,...

Smart homes for the aging population have recently started attracting attention of research community. One problems interest is this monitoring activities daily living (ADLs) elderly, in order to help identify critical problems, aiming improve their protection and general well-being. In paper, we report on our initial attempts recognize such activities, based input from networks far-field microphones distributed inside home. We propose two approaches problem: The first models entire...

In order to improve the face recognition rate under complex light conditions and address phenomenon such as white color distortion, low contrast, which might be produced while using retinex algorithm process images, this paper proposes a new multi-scale with restoration for compensation of images. On basis studying MSRCR theory, we introduce fast Fourier into original make speed faster than that traditional method. Experimental results prove contrasts, brightness, hidden details images are...