Multi-Channel Source Separation

2010 ◽  
pp. 207-245
Author(s):  
Nilesh Madhu ◽  
André Gückel
Keyword(s):  
Real Time ◽  
Source Separation ◽  
Real Life ◽  
Wide Scope ◽  
Computational Power ◽  
Challenging Problem ◽  
Signal Model ◽  
Related Field ◽  
Wide Range ◽  
Target Speaker

Machine-based multi-channel source separation in real life situations is a challenging problem, and has a wide range of applications, from medical to military. With the increase in computational power available to everyday devices, source separation in real-time has become more feasible, contributing to the boost in the research in this field in the recent past. Algorithms for source separation are based on specific assumptions regarding the source and signal model – which depends upon the application. In this chapter, the specific application considered is that of a target speaker enhancement in the presence of competing speakers and background noise. It is the aim of this contribution to present not only an exhaustive overview of state-of-the-art separation algorithms and the specific models they are based upon, but also to highlight the relations between these algorithms, where possible. Given this wide scope of the chapter, we expect it will benefit both, the student beginning his studies in the field of machine audition, and those already working in a related field and wishing to obtain an overview or insights into the field of multi-channel source separation.

scholarly journals Shadow Elimination in Soccer Game Video using Background Subtraction and Sobel Operators

2020 ◽  
pp. 2426-2433
Author(s):  
Huda Dheyauldeen Najeeb ◽  
Rana Fareed Ghani
Keyword(s):  
Real Time ◽  
Background Subtraction ◽  
Challenging Problem ◽  
Shadow Elimination ◽  
Ball Size ◽  
Soccer Game ◽  
Wide Range ◽  
Similar Appearance ◽  
New System ◽  
Sobel Operators

Object detection in real time is considered as a challenging problem. However, it is very important in a wide range of applications, especially in field of multimedia. The players and ball are the most important objects in soccer game videos and detecting them is a challenging task because of many difficulties, such as shadow and illumination, ball size, ball occluded by players or merged with lines, and similar appearance of players. To overcome these problems, we present a new system to detect the players and ball in real-time by using background subtraction and Sobel detection. The results were more accurate and approximately two times faster than those using only background subtraction.

Augmented Hardware Design of BSS in Real Time by Real Time Feasible Source Signal Separator

2020 ◽  
Vol 14 ◽  
Keyword(s):  
Real Time ◽  
Blind Source Separation ◽  
Source Separation ◽  
Separation Algorithm ◽  
Underdetermined Blind Source Separation ◽  
Memory Space ◽  
The Real ◽  
Wide Range ◽  
Source Signal ◽  
Band Limited

Blind source separation is a blooming sector in the digital signal processing for severing exact signal from the dense recorded. Exclusively, among the “Blind Source Separation” the “Under Determined Blind Source Separation” is considered than an over determined Blind Source Separation due to its wide range of usage. Nevertheless, it is seen that the real implementation is very rarely done in existing researches, because the real time implementation of UBSS (Underdetermined Blind Source Separation)is existed to be a challenging one due to its lacking hardware characteristics of increased latency, reduced speed and consumption of more memory space. Consequently, there is an increase need to implement an Underdetermined source signal separation real time with improved hardware utility that in this Unswerving framework a Real time feasible Source Signal separator is formulated in which initially the source signals are decomposed by Boosted band-limited VMD (Variational Mode Decomposition)into the “Multi component Signal” and then to an amount of "Band-Limited” IMF subjected to the Encompassed Hammer sley–Clifford source separation algorithm that uses expectation-maximization and Gibbs sampling an alternative to deterministic algorithms to determine the exact estimated parameter from E-M method. Subsequently, the source separation algorithm infers the best separation of sources signals by exact estimation and determination from the decomposed signals, whereas the iterations in E-M estimation are reduced by Gauss-Seidel method. Thus our novel source signal separator internally with a signal decomposer and a source separation algorithm with lesser number of iterations which reduces memory consumption and yields better hardware realization with reduced latency and increased speed. The proposed Implementation is done in Xilinx Platform.

A genetic algorithm for permutation flowshop scheduling under practical make-to-order production system

2016 ◽  
Vol 31 (1) ◽  
pp. 87-103 ◽  
Author(s):  
Humyun Fuad Rahman ◽  
Ruhul Sarker ◽  
Daryl Essam
Keyword(s):  
Real Time ◽  
Production System ◽  
Production Scheduling ◽  
Real Life ◽  
Real Data ◽  
Production Environment ◽  
Permutation Flowshop ◽  
Make To Order ◽  
Wide Range ◽  
Range Of Variation

AbstractThe aim of this work is to bridge the gap between the theory and actual practice of production scheduling by studying a problem from a real-life production environment. This paper considers a practical Sanitaryware production system as a number of make-to-order permutation flowshop problems. Due to the wide range of variation in its products, real-time arrival of customer orders, dynamic batch adjustments, and time for machine setup, Sanitaryware production system is complex and also time sensitive. In practice, many such companies run with suboptimal solutions. To tackle this problem, in this paper, a memetic algorithm based real-time approach has been proposed. Numerical experiments based on real data are also been presented in this paper.

scholarly journals A Multiprotocol Wireless Sensor Network for High Performance Sport Applications

2018 ◽  
Vol 1 (4) ◽  
pp. 52 ◽  
Author(s):  
Vincenzo Bonaiuto ◽  
Paolo Boatto ◽  
Nunzio Lanotte ◽  
Cristian Romagnoli ◽  
Giuseppe Annino
Keyword(s):  
Wireless Sensor Network ◽  
Real Time ◽  
Sensor Network ◽  
High Performance ◽  
Wearable Sensors ◽  
Real Life ◽  
Wireless Sensor ◽  
Sport Equipment ◽  
Wide Range ◽  
Physiological Sensors

The use of a network of wearable sensors placed on the athlete or installed into sport equipment is able to offer, in a real sport environment rather than in the unspecific spaces of a laboratory, a valuable real-time feedback to the coach during practice. This is made possible today by the coordinate use of a wide range of kinematic, dynamic, and physiological sensors. Using sensors makes training more effective, improves performance assessment, and can help in preventing injuries. In this paper, a new wireless sensor network (WSN) system for elite sport applications is presented. The network is made up of a master node and up to eight peripheral nodes (slave nodes), each one containing one or more sensors. The number of nodes can be increased with second level slave nodes; the nature of sensors varies depending on the application. Communication between nodes is made via a high performance 2.4 GHz transceiver; the network has a real-life range in excess of 100 m. The system can therefore be used in applications where the distance between nodes is long, for instance, in such sports as kayaking, sailing, and rowing. Communication with user and data download are made via a Wi-Fi link. The user communication interface is a webpage and is therefore completely platform (computer, tablet, smartphone) and operating system (Windows, iOS, Android, etc.) independent. A subset of acquired data can be visualized in real time on multiple terminals, for instance, by athlete and coach. Data from kayaking, karting, and swimming applications are presented.

scholarly journals An Academic Text Recommendation Method Based on Graph Neural Network

Information ◽  
2021 ◽  
Vol 12 (4) ◽  
pp. 172
Author(s):  
Jie Yu ◽  
Chenle Pan ◽  
Yaliu Li ◽  
Junwei Wang
Keyword(s):  
Neural Network ◽  
Real Life ◽  
Semantic Relations ◽  
Semantic Relationship ◽  
Challenging Problem ◽  
Academic Field ◽  
Wide Range ◽  
Sequential Information ◽  
Relationship Of ◽  
The Relationship

Academic text recommendation, as a kind of text recommendation, has a wide range of application prospects. Predicting texts of interest to scholars in different fields based on anonymous sessions is a challenging problem. However, the existing session-based method only considers the sequential information, and pays more attention to capture the session purpose. The relationship between adjacent items in the session is not noticed. Specifically in the field of session-based text recommendation, the most important semantic relationship of text is not fully utilized. Based on the graph neural network and attention mechanism, this paper proposes a session-based text recommendation model (TXT-SR) incorporating the semantic relations, which is applied to the academic field. TXT-SR makes full use of the tightness of semantic connections between adjacent texts. We have conducted experiments on two real-life academic datasets from CiteULike. Experimental results show that TXT-SR has better effectiveness than existing session-based recommendation methods.

scholarly journals Detecting Human and Classification of Gender using Facial Images MSIFT Features Based GSVM

2019 ◽  
Vol 8 (3) ◽  
pp. 1466-1471 ◽  
Keyword(s):  
Real Time ◽  
Real Life ◽  
Recognition System ◽  
Human Detection ◽  
Support Vector ◽  
Time Data ◽  
Processing Application ◽  
Wide Range ◽  
Human Image

Classification of gender using face recognition system is an essential concept for different types of applications in human-computer interaction and computer-aided related applications. It defines a wide range of features from human images to detect male, female and others using real-time data. There are different machine learning approaches were implemented to classify gender and also detects other images during the classification phase, which are not humans based on features extracted from human images datasets. All these existing techniques mostly depend on controlled conditions like features and other representations of the human image. Because of significant and uncertain variations of a particular image, it may be a challenging task in gender classification for real-time image processing application, whether it is male, female and others. So that in this document, we propose a Human detection and Face based gender Recognition System (HDFGR); to investigate male or female classification on real life faces using real world face databases. Our proposed approach consists Multi-Scale Invariant Feature Transform (MSIFT) to describes faces and Gaussian distance-based support vector machine (GSVM) classifier is used to classify gender and objects, i.e. male, female and other from features extracted human image datasets. We obtain an experimental performance of 98.7% by applying DSVM with boosted MSIFT features. Our proposed approach gives better classification accuracy and other performance parameters compared to different existing approaches with benchmark and evaluation of possible databases.

Ambulatory Assessment

2009 ◽  
Vol 14 (2) ◽  
pp. 109-119 ◽  
Author(s):  
Ulrich W. Ebner-Priemer ◽  
Timothy J. Trull
Keyword(s):  
Real Time ◽  
Clinical Psychology ◽  
Real Life ◽  
Ambulatory Assessment ◽  
Computer Assisted ◽  
Affective Valence ◽  
Time Data ◽  
Physiological Processes ◽  
Context Specific ◽  
Questionnaire Studies

Convergent experimental data, autobiographical studies, and investigations on daily life have all demonstrated that gathering information retrospectively is a highly dubious methodology. Retrospection is subject to multiple systematic distortions (i.e., affective valence effect, mood congruent memory effect, duration neglect; peak end rule) as it is based on (often biased) storage and recollection of memories of the original experience or the behavior that are of interest. The method of choice to circumvent these biases is the use of electronic diaries to collect self-reported symptoms, behaviors, or physiological processes in real time. Different terms have been used for this kind of methodology: ambulatory assessment, ecological momentary assessment, experience sampling method, and real-time data capture. Even though the terms differ, they have in common the use of computer-assisted methodology to assess self-reported symptoms, behaviors, or physiological processes, while the participant undergoes normal daily activities. In this review we discuss the main features and advantages of ambulatory assessment regarding clinical psychology and psychiatry: (a) the use of realtime assessment to circumvent biased recollection, (b) assessment in real life to enhance generalizability, (c) repeated assessment to investigate within person processes, (d) multimodal assessment, including psychological, physiological and behavioral data, (e) the opportunity to assess and investigate context-specific relationships, and (f) the possibility of giving feedback in real time. Using prototypic examples from the literature of clinical psychology and psychiatry, we demonstrate that ambulatory assessment can answer specific research questions better than laboratory or questionnaire studies.

Home Learning in Times of COVID: Experiences of Parents

2020 ◽  
Vol 7 (1) ◽  
pp. 9 ◽  
Author(s):  
Shelina Bhamani ◽  
Areeba Zainab Makhdoom ◽  
Vardah Bharuchi ◽  
Nasreen Ali ◽  
Sidra Kaleem ◽  
...  
Keyword(s):  
Real Life ◽  
Sampling Technique ◽  
Online Classes ◽  
Google Docs ◽  
Learning Gap ◽  
Wide Range ◽  
Home Learning ◽  
Collection Data ◽  
Learning At Home ◽  
At Home

<p align="center"><em>The widespread prevalence of COVID-19 pandemic has affected academia and parents alike. Due to the sudden closure of schools, students are missing social interaction which is vital for better learning and grooming while most schools have started online classes. This has become a tough routine for the parents working online at home since they have to ensure their children’s education. The study presented was designed to explore the experiences of home learning in times of COVID-19. A descriptive qualitative study was planned to explore the experiences of parents about home learning and management during COVID-19 to get an insight into real-life experiences.  Purposive sampling technique was used for data collection.  Data were collected from 19 parents falling in the inclusion criteria. Considering the lockdown problem, the data were collected via Google docs form with open-ended questions related to COVID-19 and home learning. Three major themes emerged after the data analysis: impact of COVID on children learning; support given by schools; and strategies used by caregivers at home to support learning. It was analyzed that the entire nation and academicians around the world have come forward to support learning at home offering a wide range of free online avenues to support parents to facilitate home-learning. Furthermore, parents too have adapted quickly to address the learning gap that have emerged in their children’s learning in these challenging times. Measures should be adopted to provide essential learning skills to children at home. Centralized data dashboards and educational technology may be used to keep the students, parents and schools updated.</em></p>

scholarly journals A Systematic Review and Qualitative Synthesis Resulting in a Typology of Elementary Classroom Movement Integration Interventions

2020 ◽  
Vol 6 (1) ◽  
Author(s):  
Spyridoula Vazou ◽  
Collin A. Webster ◽  
Gregory Stewart ◽  
Priscila Candal ◽  
Cate A. Egan ◽  
...  
Keyword(s):  
Physical Activity ◽  
Teacher Collaboration ◽  
A Priori ◽  
Real Life ◽  
Dose Intensity ◽  
Routine Practice ◽  
Qualitative Synthesis ◽  
Wide Range ◽  
Meta Analyses ◽  
Movement Integration

Abstract Background/Objective Movement integration (MI) involves infusing physical activity into normal classroom time. A wide range of MI interventions have succeeded in increasing children’s participation in physical activity. However, no previous research has attempted to unpack the various MI intervention approaches. Therefore, this study aimed to systematically review, qualitatively analyze, and develop a typology of MI interventions conducted in primary/elementary school settings. Subjects/Methods Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed to identify published MI interventions. Irrelevant records were removed first by title, then by abstract, and finally by full texts of articles, resulting in 72 studies being retained for qualitative analysis. A deductive approach, using previous MI research as an a priori analytic framework, alongside inductive techniques were used to analyze the data. Results Four types of MI interventions were identified and labeled based on their design: student-driven, teacher-driven, researcher-teacher collaboration, and researcher-driven. Each type was further refined based on the MI strategies (movement breaks, active lessons, other: opening activity, transitions, reward, awareness), the level of intrapersonal and institutional support (training, resources), and the delivery (dose, intensity, type, fidelity). Nearly half of the interventions were researcher-driven, which may undermine the sustainability of MI as a routine practice by teachers in schools. An imbalance is evident on the MI strategies, with transitions, opening and awareness activities, and rewards being limitedly studied. Delivery should be further examined with a strong focus on reporting fidelity. Conclusions There are distinct approaches that are most often employed to promote the use of MI and these approaches may often lack a minimum standard for reporting MI intervention details. This typology may be useful to effectively translate the evidence into practice in real-life settings to better understand and study MI interventions.

scholarly journals DOLARS, a Distributed On-Line Activity Recognition System by Means of Heterogeneous Sensors in Real-Life Deployments—A Case Study in the Smart Lab of The University of Almería

Sensors ◽  
2021 ◽  
Vol 21 (2) ◽  
pp. 405
Author(s):  
Marcos Lupión ◽  
Javier Medina-Quero ◽  
Juan F. Sanjuan ◽  
Pilar M. Ortigosa
Keyword(s):  
Real Time ◽  
Activity Recognition ◽  
Real Life ◽  
Recognition System ◽  
Machine Learning Algorithms ◽  
Sensor Data ◽  
Heterogeneous Sensors ◽  
On Line ◽  
The University

Activity Recognition (AR) is an active research topic focused on detecting human actions and behaviours in smart environments. In this work, we present the on-line activity recognition platform DOLARS (Distributed On-line Activity Recognition System) where data from heterogeneous sensors are evaluated in real time, including binary, wearable and location sensors. Different descriptors and metrics from the heterogeneous sensor data are integrated in a common feature vector whose extraction is developed by a sliding window approach under real-time conditions. DOLARS provides a distributed architecture where: (i) stages for processing data in AR are deployed in distributed nodes, (ii) temporal cache modules compute metrics which aggregate sensor data for computing feature vectors in an efficient way; (iii) publish-subscribe models are integrated both to spread data from sensors and orchestrate the nodes (communication and replication) for computing AR and (iv) machine learning algorithms are used to classify and recognize the activities. A successful case study of daily activities recognition developed in the Smart Lab of The University of Almería (UAL) is presented in this paper. Results present an encouraging performance in recognition of sequences of activities and show the need for distributed architectures to achieve real time recognition.

Sign in / Sign up

Export Citation Format

Share Document