This thesis investigates an implementation of speech recognition front-end. It is an application specific integrated circuit (ASIC) solution. A Mel Cepstrum algorithm is implemented for the feature extraction. We present a new mixed split-radix and radix-2 Fast Fourier Transform (FFT) algorithm, which can effectively minimize the number of complex multiplications in the speech recognition front-end. A prime length discrete cosine transform (DCT) is done effectively through the use of two shorter length correlations. The algorithm results in a circular correlation structure that is suitable for a constant coefficient multiplication and shift-register realization. The multiplicative normalization algorithm is used for square root function. Radix-2 algorithm is used in the first 5 stages and radix-4 algorithm is used in the other stages to speed up the convergence. A similar normalization algorithm is present for natural logarithm. / Graduation date: 2002

Fördomar och diskriminering existerar ännu i dagens samhälle. Rasfördomar förekommer dagligen i olika arbetsmiljöer och även under anställningsprocesser. Denna undersökning syftar till att studera och utforska om namn spelar en avgörande roll vid bedömning av en arbetsansökan. För att besvara syftet med denna undersökning utfördes en undersökning där studenter vid Mälardalens Högskola fick läsa en arbetsannons och en arbetsansökan med olika namn. Därefter fick deltagarna skatta och bedöma de arbetssökande utifrån tolv egenskaper samt finna om de sökande var lämpliga för arbetstjänsten som civilingenjör. Undersökningen påvisade att namn inte har betydelse för hur man bedömer en arbetssökande då det inte fanns en signifkant skillnad mellan namn och hur man bedömde en arbetssökande.

A recent development in drift-control for agricultural ground sprayers is the installation of a hood (or shroud) over the boom. Hoods are designed to provide a protected zone in which droplets can be sprayed and deposited on the target with limited interference from the wind. This study compared chemical drift using hooded and open-boom pesticide sprayers in various configurations. The hooded sprayer included an air-foil mounted on top of the hood, designed to re-direct the airflow and further decrease drift. The principal objective was to test the performance of the hood, the air-foil, and nozzle size in reducing drift. A fluorescent dye and water solution was sprayed adjacent to a series of parallel string collectors. The amount of drifting material was determined by rinsing the string and testing the fluorescence of the rinsewater. Wind speed and direction, temperature, and relative humidity were monitored during each field test so that the influence of weather conditions on drift could be assessed. Statistical and graphical comparisons were based on the development of a regression model to describe downwind drift for each sprayer configuration under a set of variable weather conditions. Only parameters found to be important to drift were included in the final model. These parameters included sprayer configuration, wind speed, temperature, and the interactions of configuration with wind speed and temperature. Results of the comparisons indicated that for a larger droplet spectrum, the hood reduced drift significantly, while for smaller droplets the hood was ineffective. The air-foil was found to have no influence on drift. / Graduation date: 1991

With widespread use of pesticides in modern agriculture, the impacts of spray drift have become a topic of considerable interest. The drifting of sprays is a highly complex process influenced by many factors. Advances in aerial application technology and in our ability to measure drift, coupled with the adoption of new technologies for regulating pesticide application have necessitated further research in the pesticide application process. Experiments were conducted to quantify spray drift and describe its movement from aerial applications of pesticide. The effects of spray quality, atomizer type and ground cover were examined. Initial airborne drift amounts were greater than downwind deposits, thus not all of the drifting spray was deposited in the measuring area. Total off-target movement of spray was significantly greater for Fine compared to Coarse sprays. Rotary and hydraulic atomizers, both producing Fine sprays, produced statistically similar off-target movement of sprays. Similarly, no significant statistical differences in spray drift between applications to bare ground and applications to a headed barley crop canopy were not identified. Contrary to expectations, aerial application to bare ground seemed to result in less off-target movement than application to a crop canopy. The vertical spray cloud profiles were similar for all applications with the greatest amount of spray present at the height of release. Spray concentrations diminished from that height upward with diffusion and downward with deposition. The empirical data disagreed with the mechanistic model AgDISP which is currently used in the Canadian regulatory process. The model over-predicted drift deposition by a factor of two to five. Variability in spray deposit values could not be attributed to average differences in meteorological conditions at the time of application. Experiments with appropriate protocols for increased sensitivity may be required to more accurately report subtle differences in drift at distances greater than 200 m from the target area.

With the advance in applying technology in education, the traditional lecture-driven teaching style is gradually replaced by a more active teaching style where the students play a more active rule in the learning process. In this paper we introduce a new initiative to provide a suite of online tools for learning DEVS model. The uniqueness of this tutorial project is the integration of information technology and multimedia into education through the development of an interactive tutorial and the characteristic of contextualized learning. The tutorial teaches students about the basic aspects of discrete event system and simulation. The interactive tutorial fully utilizes the power of the information and multimedia technology, web application and the programming language Java, to enhance students’ learning to achieve rich interactivity. The tutorial greatly supports human-computer collaboration to enhance learning and to satisfy user goals by effectively allowing the user to interact.

In this thesis we explore several algorithms for automatic real-time tracking of a tennis ball. We first investigate the use of background subtraction with color/shape recognition for fast tracking of the tennis ball. We then compare our solution with a cascade of boosted Haar classifiers [68] in a simulated environment to estimate the accuracy and ideal processing speeds. The results show that background subtraction techniques were not only faster but also more accurate than Haar classifiers. Following these promising results, we extend the background subtraction and develop other three improved techniques. These techniques use more accurate background models, more reliable and stringent criteria. They allow us to track the tennis ball in a real tennis environment with cameras having higher resolutions and frame rates. <p>We tested our techniques with a large number of real tennis videos. In the indoors environment, We achieved a true positive rate of about 90%, a false alarm rate of less than 2%, and a tracking speed of about 20 fps. For the outdoors environment, the performance of our techniques is not as good as the indoors cases due to the complexity and instability of the outdoors environment. The problem can be solved by resetting our system such that the camera focuses mainly on the tennis ball. Therefore, the influence of the external factors is minimized.<p>Despite the existing limitations, our techniques are able to track a tennis ball with very high accuracy and fast speed which can not be achieved by most tracking techniques currently available. We are confident that the motion information generated from our techniques is reliable and accurate. Giving this promising result, we believe some real-world applications can be constructed.

La fibroïne de soie extraite des fibres de cocons de vers à soie Bombyx mori a été utilisée dans une vaste gamme d'applications telles que la culture cellulaire, la régénération de tissus osseux et la libération contrôlée de molécules actives. Le choix de cette protéine fibreuse pour la conception de biomatériaux voués à l'ingénierie tissulaire est motivé par leurs impressionnantes propriétés mécaniques, leur biocompatibilité et leur biodégradabilité. Le but du projet de recherche principal présenté dans ce mémoire était de préparer des matériaux poreux à base de fibroïne de soie séricicole pour créer une matrice 3D utilisable comme support pour culture cellulaire in vitro. Dans cette optique, l'impact de l'addition de NaCl (0-250 mM) dans les solutions de fibroïne avant la congélation ainsi que l'effet de la température de congélation (-22°C ou -73°C) ont été étudiés lors de la préparation de matrices 3D lyophilisées. Les matériaux ont ensuite été traités au méthanol pour induire leur insolubilité, puis immergés dans l'eau pour extraire le sel. Tel qu'observé par microscopie électronique à balayage (MEB), la morphologie des matrices préparées à -22°C est constituée d'un réseau poreux interconnecté devenant moins serré avec l'ajout de sel. Les biomatériaux préparés à -73°C possèdent plutôt une structure en feuillets orientés sur lesquels des filaments et micropores sont apparus lors de l'addition de sel. Aux deux températures de congélation, les matrices ont affiché un ratio de gonflement plus élevé et une résistance à la compression plus faible lors de leur préparation en présence de sel. Les différences morphologiques des matrices ont été corrélées avec les variations dans les processus de nucléation et de croissance des cristaux de glace lors de la congélation. L'étude des matériaux par spectroscopie infrarouge à transformée de Fourier (FTIR) et diffraction des rayons X (XRD) a révélé un changement dans la structure secondaire de la fibroïne, passant d'un mélange d'hélices a et pelotes statistiques dynamiques, vers une structure majoritairement en feuillets β induite par le traitement au méthanol. Aucune différence n'a cependant été observée en fonction de l'ajout de sel ou de la variation de la température de congélation. La microscopie confocale à balayage laser a permis de constater que des cellules ensemencées dans les matrices demeuraient viables et avaient pénétré la majorité des biomatériaux pour s'organiser en agrégats après 24 h de culture. Cette étude permet de conclure qu'il est possible d'adapter la morphologie de matrices 3D lyophilisées pour une utilisation comme support pour culture cellulaire. Aussi, un second projet de recherche présenté dans ce mémoire aura permis d'évaluer le potentiel d'utilisation d'un isolat de protéines de soja mélangé à l'amidon riche en amylose (ARA) pour la production d'un excipient pharmaceutique visant la libération de molécules actives par voie orale. Les résultats de libération de la 8-hydroxyquinoline (HQN) comme traceur ont montré que l'ajout d'ARA gélatinisé n'améliore pas le comportement de l'isolat de protéines de soja Pro, cependant l'ajout de Pro semble plutôt améliorer les propriétés de l'ARA gélatinisé comme excipient. La Pro ajoutée aux formulations d'ARA a apparemment favorisé la cohésion des particules en milieux digestifs simulés, présentant un potentiel d'additif liant dans les formulations. ______________________________________________________________________________ MOTS-CLÉS DE L’AUTEUR : fibroïne de soie, matrices 3D, culture cellulaire, lyophilisation, cristaux de glace, chlorure de sodium, température de congélation, isolat de protéines de soja, amidon

L'environnement de départ de ce projet était le site Web SmartShopping, un portail permettant de naviguer à travers les différents spéciaux quotidiens des magasins d'alimentation de l'île de Montréal, puis de les ajouter à un panier, et enfin d'observer le trajet nécessaire afin de visiter les différents magasins d'où proviennent ces spéciaux. Le but du projet était d'implémenter l'affichage d'un trajet optimal de type Voyageur de Commerce Généralisé entre les différentes franchises des enseignes à visiter, sur une carte GoogleMapsTM, puis incorporer cette fonctionnalité au site Web SmartShopping. L'algorithme précédemment en place choisissait, pour établir un trajet, les magasins qui se trouvaient les plus proches du point de départ, soit l'adresse du client, pour chaque enseigne à visiter. Ce travail consistait donc à comparer les algorithmes de pointe du moment afin d'implémenter le meilleur d'entre eux en termes de rapidité et d'optimalité, pour un échantillon de petite taille. Après analyse et comparaison, un algorithme de type génétique créé par Tasgetiren et al. [TSPL07] a été implémenté en langage C++, en relation avec une page Web codée en PHP, et avec transmission des paramètres par fichiers texte. Par rapport à l'ancien algorithme, les résultats de ce travail montrent une nette amélioration des trajets proposés, et ceci dans l'ensemble des cas testés, avec une moyenne de baisse des distances de 12%, pour les cas de 2 à 5 magasins. Le site Web avec sa nouvelle fonctionnalité peut être consulté à l'adresse URL suivante : http://www.trex.uqam.ca/~smartshopping. ______________________________________________________________________________ MOTS-CLÉS DE L’AUTEUR : algorithme, algorithme de voyageur de commerce généralisé, trajet optimal, site web

Researchers are in broad agreement that energy-conserving actions produce economic as well as energy savings. Household energy rating systems (HERS) have been established in many countries to inform households of their house’s current energy performance and to help reduce their energy consumption and greenhouse gas emissions. In Canada, the national EnerGuide for Houses (EGH) program is delivered by many local delivery agents, including non-profit green community organizations. Waterloo Region Green Solutions is the local non-profit that offers the EGH residential energy evaluation service to local households. The purpose of this thesis is to explore the determinants of household’s participation in the residential energy efficiency program (REEP) in Waterloo Region, to explain the relationship between the explanatory variables and REEP participation, and to propose ways to improve this kind of program. A spatial (trend) analysis was conducted within a geographic information system (GIS) to determine the spatial patterns of the REEP participation in Waterloo Region from 1999 to 2006. The impact of sources of information on participation and relationships between participation rates and explanatory variables were identified. GIS proved successful in presenting a visual interpretation of spatial patterns of the REEP participation. In general, the participating households tend to be clustered in urban areas and scattered in rural areas. Different sources of information played significant roles in reaching participants in different years. Moreover, there was a relationship between each explanatory variable and the REEP participation rates. Statistical analysis was applied to obtain a quantitative assessment of relationships between hypothesized explanatory variables and participation in the REEP. The Poisson regression model was used to determine the relationship between hypothesized explanatory variables and REEP participation at the CDA level. The results show that all of the independent variables have a statistically significant positive relationship with REEP participation. These variables include level of education, average household income, employment rate, home ownership, population aged 65 and over, age of home, and number of eligible dwellings. The logistic regression model was used to assess the ability of the hypothesized explanatory variables to predict whether or not households would participate in a second follow-up evaluation after completing upgrades to their home. The results show all the explanatory variables have significant relationships with the dependent variable. The increased rating score, average household income, aged population, and age of home are positively related to the dependent variable. While the dwelling size and education has negative relationships with the dependent variable. In general, the contribution of this work provides a practical understanding of how the energy efficiency program operates, and insight into the type of variables that may be successful in bringing about changes in performance in the energy efficiency project in Waterloo Region. Secondly, with the completion of this research, future residential energy efficiency programs can use the information from this research and emulate or expand upon the efforts and lessons learned from the Residential Energy Efficiency Project in Waterloo Region case study. Thirdly, this research also contributes to practical experience on how to integrate different datasets using GIS.

This thesis has been divided into two essential parts, the purpose of the first part is to investigate and explore a three-tier architecture for remote health monitoring system capable to collect, store and forward the physiological data, which has been collected by a mobile device via a bluetooth connection from body sensors, to an internet data base server. During the second part, we have tried to take a deep look into a heart beat modeling method. We have studied and investigated on extended integral pulse frequency modulation model which is used for the presence of ectopic beats and heart rate turbulence.