Generative Part-Based Gabor Object Detector

Object detection is a fundamental task of computer vision that is utilized as a core part in a number of industrial and scientific applications, for example, in robotics, where objects need to be correctly detected and localized prior to being grasped and manipulated. Existing object detectors vary in (i) the amount of supervision they need for training, (ii) the type of a learning method adopted (generative or discriminative) and (iii) the amount of spatial information used in the object model (model-free, using no spatial information in the object model, or model-based, with the explicit spatial model of an object). Although some existing methods report good performance in the detection of certain objects, the results tend to be application specific and no universal method has been found that clearly outperforms all others in all areas. This work proposes a novel generative part-based object detector. The generative learning procedure of the developed method allows learning from positive examples only. The detector is based on finding semantically meaningful parts of the object (i.e. a part detector) that can provide additional information to object location, for example, pose. The object class model, i.e. the appearance of the object parts and their spatial variance, constellation, is explicitly modelled in a fully probabilistic manner. The appearance is based on bio-inspired complex-valued Gabor features that are transformed to part probabilities by an unsupervised Gaussian Mixture Model (GMM). The proposed novel randomized GMM enables learning from only a few training examples. The probabilistic spatial model of the part configurations is constructed with a mixture of 2D Gaussians. The appearance of the parts of the object is learned in an object canonical space that removes geometric variations from the part appearance model. Robustness to pose variations is achieved by object pose quantization, which is more efficient than previously used scale and orientation shifts in the Gabor feature space. Performance of the resulting generative object detector is characterized by high recall with low precision, i.e. the generative detector produces large number of false positive detections. Thus a discriminative classifier is used to prune false positive candidate detections produced by the generative detector improving its precision while keeping high recall. Using only a small number of positive examples, the developed object detector performs comparably to state-of-the-art discriminative methods.

Visual Customer Segmentation and Data Exploration with the Self-Organizing Map

Advancements in information technology have made it possible for organizations to gather and store vast amounts of data of their customers. Information stored in databases can be highly valuable for organizations. However, analyzing large databases has proven to be difficult in practice. For companies in the retail industry, customer intelligence can be used to identify profitable customers, their characteristics, and behavior. By clustering customers into homogeneous groups, companies can more effectively manage their customer base and target profitable customer segments. This thesis will study the use of the self-organizing map (SOM) as a method for analyzing large customer datasets, clustering customers, and discovering information about customer behavior. Aim of the thesis is to find out whether the SOM could be a practical tool for retail companies to analyze their customer data.

The infared optical properties of Sr2RuO4 and SmTiO3 including an object-oriented resistivity interface /

The perovskite crystal structure is host to many different materials from insulating to superconducting providing a diverse range of intrinsic character and complexity. A better fundamental description of these materials in terms of their electronic, optical and magnetic properties undoubtedly precedes an effective realization of their application potential. SmTiOa, a distorted perovskite has a strongly localized electronic structure and undergoes an antiferromagnetic transition at 50 K in its nominally stoichiometric form. Sr2Ru04 is a layered perovskite superconductor (ie. Tc % 1 K) bearing the same structure as the high-tem|>erature superconductor La2_xSrrCu04. Polarized reflectance measurements were carried out on both of these materials revealing several interesting features in the far-infrared range of the spectrum. In the case of SmTiOa, although insulating, evidence indicates the presence of a finite background optical conductivity. As the temperature is lowered through the ordering temperature a resonance feature appears to narrow and strengthen near 120 cm~^ A nearby phonon mode appears to also couple to this magnetic transition as revealed by a growing asymmetry in the optica] conductivity. Experiments on a doped sample with a greater itinerant character and lower Neel temperature = 40 K also indicate the presence of this strongly temperature dependent mode even at twice the ordering temperature. Although the mode appears to be sensitive to the magnetic transition it is unclear whether a magnon assignment is appropriate. At very least, evidence suggests an interesting interaction between magnetic and electronic excitations. Although Sr2Ru04 is highly anisotropic it is metallic in three-dimensions at low temperatures and reveals its coherent transport in an inter-plane Drude-like component to the highest temperatures measured (ie. 90 K). An extended Drude analysis is used to probe the frequency dependent scattering character revealing a peak in both the mass enhancement and scattering rate near 80 cm~* and 100 cm~* respectively. All of these experimental observations appear relatively consistent with a Fermi-liquid picture of charge transport. To supplement the optical measurements a resistivity station was set up with an event driven object oriented user interface. The program controls a Keithley Current Source, HP Nano-Voltmeter and Switching Unit as well as a LakeShore Temperature Controller in order to obtain a plot of the Resistivity as a function of temperature. The system allows for resistivity measurements ranging from 4 K to 290 K using an external probe or between 0.4 K to 295 K using a Helium - 3 Cryostat. Several materials of known resistivity have confirmed the system to be robust and capable of measuring metallic samples distinguishing features of several fiQ-cm.

Winâ Foy : functional object-oriented programming language

This thesis will introduce a new strongly typed programming language utilizing Self types, named Win--*Foy, along with a suitable user interface designed specifically to highlight language features. The need for such a programming language is based on deficiencies found in programming languages that support both Self types and subtyping. Subtyping is a concept that is taken for granted by most software engineers programming in object-oriented languages. Subtyping supports subsumption but it does not support the inheritance of binary methods. Binary methods contain an argument of type Self, the same type as the object itself, in a contravariant position, i.e. as a parameter. There are several arguments in favour of introducing Self types into a programming language (11. This rationale led to the development of a relation that has become known as matching [4, 5). The matching relation does not support subsumption, however, it does support the inheritance of binary methods. Two forms of matching have been proposed (lJ. Specifically, these relations are known as higher-order matching and I-bound matching. Previous research on these relations indicates that the higher-order matching relation is both reflexive and transitive whereas the f-bound matching is reflexive but not transitive (7]. The higher-order matching relation provides significant flexibility regarding inheritance of methods that utilize or return values of the same type. This flexibility, in certain situations, can restrict the programmer from defining specific classes and methods which are based on constant values [21J. For this reason, the type This is used as a second reference to the type of the object that cannot, contrary to Self, be specialized in subclasses. F-bound matching allows a programmer to define a function that will work for all types of A', a subtype of an upper bound function of type A, with the result type being dependent on A'. The use of parametric polymorphism in f-bound matching provides a connection to subtyping in object-oriented languages. This thesis will contain two main sections. Firstly, significant details concerning deficiencies of the subtype relation and the need to introduce higher-order and f-bound matching relations into programming languages will be explored. Secondly, a new programming language named Win--*Foy Functional Object-Oriented Programming Language has been created, along with a suitable user interface, in order to facilitate experimentation by programmers regarding the matching relation. The construction of the programming language and the user interface will be explained in detail.

Equational Reasoning about Object-Oriented Programs

Formal verification of software can be an enormous task. This fact brought some software engineers to claim that formal verification is not feasible in practice. One possible method of supporting the verification process is a programming language that provides powerful abstraction mechanisms combined with intensive reuse of code. In this thesis we present a strongly typed functional object-oriented programming language. This language features type operators of arbitrary kind corresponding to so-called type protocols. Sub classing and inheritance is based on higher-order matching, i.e., utilizes type protocols as basic tool for reuse of code. We define the operational and axiomatic semantics of this language formally. The latter is the basis of the interactive proof assistant VOOP (Verified Object-Oriented Programs) that allows the user to prove equational properties of programs interactively.

Real-Time Automatic Object Classification and Tracking using Genetic Programming and NVIDIA R CUDA TM

Genetic Programming (GP) is a widely used methodology for solving various computational problems. GP's problem solving ability is usually hindered by its long execution times. In this thesis, GP is applied toward real-time computer vision. In particular, object classification and tracking using a parallel GP system is discussed. First, a study of suitable GP languages for object classification is presented. Two main GP approaches for visual pattern classification, namely the block-classifiers and the pixel-classifiers, were studied. Results showed that the pixel-classifiers generally performed better. Using these results, a suitable language was selected for the real-time implementation. Synthetic video data was used in the experiments. The goal of the experiments was to evolve a unique classifier for each texture pattern that existed in the video. The experiments revealed that the system was capable of correctly tracking the textures in the video. The performance of the system was on-par with real-time requirements.

Object-Oriented Genetic Programming for the Automatic Inference of Graph Models for Complex Networks

Complex networks are systems of entities that are interconnected through meaningful relationships. The result of the relations between entities forms a structure that has a statistical complexity that is not formed by random chance. In the study of complex networks, many graph models have been proposed to model the behaviours observed. However, constructing graph models manually is tedious and problematic. Many of the models proposed in the literature have been cited as having inaccuracies with respect to the complex networks they represent. However, recently, an approach that automates the inference of graph models was proposed by Bailey [10] The proposed methodology employs genetic programming (GP) to produce graph models that approximate various properties of an exemplary graph of a targeted complex network. However, there is a great deal already known about complex networks, in general, and often specific knowledge is held about the network being modelled. The knowledge, albeit incomplete, is important in constructing a graph model. However it is difficult to incorporate such knowledge using existing GP techniques. Thus, this thesis proposes a novel GP system which can incorporate incomplete expert knowledge that assists in the evolution of a graph model. Inspired by existing graph models, an abstract graph model was developed to serve as an embryo for inferring graph models of some complex networks. The GP system and abstract model were used to reproduce well-known graph models. The results indicated that the system was able to evolve models that produced networks that had structural similarities to the networks generated by the respective target models.

Segmentation d’images intravasculaires ultrasonores

L'imagerie intravasculaire ultrasonore (IVUS) est une technologie médicale par cathéter qui produit des images de coupe des vaisseaux sanguins. Elle permet de quantifier et d'étudier la morphologie de plaques d'athérosclérose en plus de visualiser la structure des vaisseaux sanguins (lumière, intima, plaque, média et adventice) en trois dimensions. Depuis quelques années, cette méthode d'imagerie est devenue un outil de choix en recherche aussi bien qu'en clinique pour l'étude de la maladie athérosclérotique. L'imagerie IVUS est par contre affectée par des artéfacts associés aux caractéristiques des capteurs ultrasonores, par la présence de cônes d'ombre causés par les calcifications ou des artères collatérales, par des plaques dont le rendu est hétérogène ou par le chatoiement ultrasonore (speckle) sanguin. L'analyse automatisée de séquences IVUS de grande taille représente donc un défi important. Une méthode de segmentation en trois dimensions (3D) basée sur l'algorithme du fast-marching à interfaces multiples est présentée. La segmentation utilise des attributs des régions et contours des images IVUS. En effet, une nouvelle fonction de vitesse de propagation des interfaces combinant les fonctions de densité de probabilité des tons de gris des composants de la paroi vasculaire et le gradient des intensités est proposée. La segmentation est grandement automatisée puisque la lumière du vaisseau est détectée de façon entièrement automatique. Dans une procédure d'initialisation originale, un minimum d'interactions est nécessaire lorsque les contours initiaux de la paroi externe du vaisseau calculés automatiquement sont proposés à l'utilisateur pour acceptation ou correction sur un nombre limité d'images de coupe longitudinale. La segmentation a été validée à l'aide de séquences IVUS in vivo provenant d'artères fémorales provenant de différents sous-groupes d'acquisitions, c'est-à-dire pré-angioplastie par ballon, post-intervention et à un examen de contrôle 1 an suivant l'intervention. Les résultats ont été comparés avec des contours étalons tracés manuellement par différents experts en analyse d'images IVUS. Les contours de la lumière et de la paroi externe du vaisseau détectés selon la méthode du fast-marching sont en accord avec les tracés manuels des experts puisque les mesures d'aire sont similaires et les différences point-à-point entre les contours sont faibles. De plus, la segmentation par fast-marching 3D s'est effectuée en un temps grandement réduit comparativement à l'analyse manuelle. Il s'agit de la première étude rapportée dans la littérature qui évalue la performance de la segmentation sur différents types d'acquisition IVUS. En conclusion, la segmentation par fast-marching combinant les informations des distributions de tons de gris et du gradient des intensités des images est précise et efficace pour l'analyse de séquences IVUS de grandes tailles. Un outil de segmentation robuste pourrait devenir largement répandu pour la tâche ardue et fastidieuse qu'est l'analyse de ce type d'images.

Patterns and quality of object-oriented software systems

Lors de ces dix dernières années, le coût de la maintenance des systèmes orientés objets s'est accru jusqu' à compter pour plus de 70% du coût total des systèmes. Cette situation est due à plusieurs facteurs, parmi lesquels les plus importants sont: l'imprécision des spécifications des utilisateurs, l'environnement d'exécution changeant rapidement et la mauvaise qualité interne des systèmes. Parmi tous ces facteurs, le seul sur lequel nous ayons un réel contrôle est la qualité interne des systèmes. De nombreux modèles de qualité ont été proposés dans la littérature pour contribuer à contrôler la qualité. Cependant, la plupart de ces modèles utilisent des métriques de classes (nombre de méthodes d'une classe par exemple) ou des métriques de relations entre classes (couplage entre deux classes par exemple) pour mesurer les attributs internes des systèmes. Pourtant, la qualité des systèmes par objets ne dépend pas uniquement de la structure de leurs classes et que mesurent les métriques, mais aussi de la façon dont celles-ci sont organisées, c'est-à-dire de leur conception, qui se manifeste généralement à travers les patrons de conception et les anti-patrons. Dans cette thèse nous proposons la méthode DEQUALITE, qui permet de construire systématiquement des modèles de qualité prenant en compte non seulement les attributs internes des systèmes (grâce aux métriques), mais aussi leur conception (grâce aux patrons de conception et anti-patrons). Cette méthode utilise une approche par apprentissage basée sur les réseaux bayésiens et s'appuie sur les résultats d'une série d'expériences portant sur l'évaluation de l'impact des patrons de conception et des anti-patrons sur la qualité des systèmes. Ces expériences réalisées sur 9 grands systèmes libres orientés objet nous permettent de formuler les conclusions suivantes: • Contre l'intuition, les patrons de conception n'améliorent pas toujours la qualité des systèmes; les implantations très couplées de patrons de conception par exemple affectent la structure des classes et ont un impact négatif sur leur propension aux changements et aux fautes. • Les classes participantes dans des anti-atrons sont beaucoup plus susceptibles de changer et d'être impliquées dans des corrections de fautes que les autres classes d'un système. • Un pourcentage non négligeable de classes sont impliquées simultanément dans des patrons de conception et dans des anti-patrons. Les patrons de conception ont un effet positif en ce sens qu'ils atténuent les anti-patrons. Nous appliquons et validons notre méthode sur trois systèmes libres orientés objet afin de démontrer l'apport de la conception des systèmes dans l'évaluation de la qualité.

Segmentation et exclusion des Gitans de Perpignan : émergence d'une élite politique?

Dans la dernière décennie, une abondante littérature a documenté la situation des Rroms d'Europe Centrale et de l'Est, où a émergé une nouvelle élite politiquement activiste. Mais chez les Tsiganes d'Europe de l'Ouest, l’activisme politique d’une élite semblait absent. Cette étude de terrain a été réalisée chez Gitans de Perpignan, à la recherche d’une action et d’une élite politique chez ce groupe, dans le contexte culturel d’une société segmentaire à pouvoir diffus, frappée d’exclusion par la société majoritaire. En effet, je propose que le concept de société segmentaire puisse s’appliquer aux Gitans, et que l’exclusion des Gitans par les païos (non Gitans) constitue un déni de la réalité relationnelle des Gitans avec la majorité païa. Enfin, l’enquête a révélé la position de «médiateurs culturels» des différents agents qui interviennent entre le monde des Gitans et celui des païos. C’est à travers le rôle de «médiateurs culturels» qu’émerge peut-être une élite politique.