Rafik Goubran

Bio: Rafik Goubran is an academic researcher from Carleton University. The author has contributed to research in topics: Noise & Adaptive filter. The author has an hindex of 35, co-authored 344 publications receiving 4866 citations. Previous affiliations of Rafik Goubran include Polytechnic University of Catalonia & Nortel.

Robust voice activity detection using higher-order statistics in the LPC residual domain

Abstract: This paper presents a robust algorithm for voice activity detection (VAD) based on newly established properties of the higher order statistics (HOS) of speech. Analytical expressions for the third and fourth-order cumulants of the LPC residual of short-term speech are derived assuming a sinusoidal model. The flat spectral feature of this residual results in distinct characteristics for these cumulants in terms of phase, periodicity and harmonic content and yields closed-form expressions for the skewness and kurtosis. Important properties about these cumulants and their similarity with the autocorrelation function are revealed from this exploratory part. They show that the HOS of speech are sufficiently distinct from those of Gaussian noise and can be used as a basis for speech detection. Their immunity to Gaussian noise makes them particularly useful in algorithms designed for low SNR environments. The proposed VAD algorithm combines HOS metrics with second-order measures, such as SNR and LPC prediction error, to classify speech and noise frames. The variance of the HOS estimators is quantified and used to yield a likelihood measure for noise frames. Moreover, a voicing condition for speech frames is derived based on the relation between the skewness and kurtosis of voiced speech. The performance of the algorithm is compared to the ITU-T G.729B VAD in various noise conditions, and quantified using the probability of correct and false classifications. The results show that the proposed algorithm has an overall better performance than G.729B, with noticeable improvement in Gaussian-like noises, such as street and parking garage, and moderate to low SNR.

Integration of Smart Home Technologies in a Health Monitoring System for the Elderly

Abstract: Among older adults, the challenges of maintaining mobility and cognitive function make it increasingly difficult to remain living alone independently. As a result, many older adults are forced to seek residence in costly clinical institutions where they can receive constant medical supervision. A home-based automated system that monitors their health and well- being while remaining unobtrusive would provide them with a more comfortable and independent lifestyle, as well as more affordable care. This paper presents a smart home system for the elderly, developed by the Technology Assisted Friendly Environment for the Third Age (TAFETA) group. It introduces the sensor technologies integrated in the system and develops a framework for the processing and communication of the extracted information. It also considers the acceptability and implications of this technology from the perspective of the potential occupants.

Speech quality prediction in VoIP using the extended E-model

Abstract: The paper investigates the effects of packet loss and delay jitter on speech quality in voice over Internet protocol (VoIP) scenarios. A new formula is proposed to quantify these effects and incorporated into ITU-T G.107, the E-model. In the simulation, codecs ITU-T G.723.1 and G.729 are used; random packet loss and Pareto distributed network delay are introduced. The prediction errors range between -0.20 and +0.12 MOS (mean opinion score). The formula extends the coverage of the current E-model, and is very useful in MOS prediction as well as network planning.

Privacy versus autonomy: A tradeoff model for smart home monitoring technologies

Abstract: Smart homes are proposed as a new location for the delivery of healthcare services. They provide healthcare monitoring and communication services, by using integrated sensor network technologies. We validate a hypothesis regarding older adults' adoption of home monitoring technologies by conducting a literature review of articles studying older adults' attitudes and perceptions of sensor technologies. Using current literature to support the hypothesis, this paper applies the tradeoff model to decisions about sensor acceptance. Older adults are willing to trade privacy (by accepting a monitoring technology), for autonomy. As the information captured by the sensor becomes more intrusive and the infringement on privacy increases, sensors are accepted if the loss in privacy is traded for autonomy. Even video cameras, the most intrusive sensor type were accepted in exchange for the height of autonomy which is to remain in the home.

Feature extraction for the differentiation of dry and wet cough sounds

Abstract: Differentiating dry and wet cough is an important factor in respiratory disease. The main objective of this paper is to analyze cough sounds and extract features that can be used in differentiation of dry and wet cough sounds. This paper proposes two features to achieve this goal. The first feature is the number of peaks of the energy envelope of the cough signal. The second feature is the power ratio of two frequency bands of the second phase of the cough signal. A set of eight highly dry and eight highly wet cough sound recordings were used. Using these two features, a clear separation was observed among the dry and wet cough sound recordings.

An Introduction To The Event Related Potential Technique

Abstract: Thank you for downloading an introduction to the event related potential technique. Maybe you have knowledge that, people have look hundreds times for their favorite readings like this an introduction to the event related potential technique, but end up in malicious downloads. Rather than reading a good book with a cup of tea in the afternoon, instead they are facing with some malicious bugs inside their laptop.

Instrumentation for trace detection of high explosives

Abstract: There is at present an urgent need for trace detection of high explosives, with applications to screening of people, packages, luggage, and vehicles. A great concern, because of recent terrorist activities, is for the development of methods that might allow detection and identification of explosives at a stand off distance. Nearly every analytical chemical method has been or is being applied to this problem. This review outlines the properties of explosives that might be utilized in detection schemes, discusses sampling issues, presents recent method developments with particular attention to detection limits, speed of analysis, and portability, and looks towards future developments.

Recurrent Neural Networks for Prediction

Abstract: Within this text neural networks are considered as massively interconnected nonlinear adaptive filters.