Made available in DSpace on 2014-12-17T14:54:56Z (GMT). No. of bitstreams: 1 NathaleeCA_DISSERT.pdf: 1318151 bytes, checksum: d2471205a640d8428567d06ace6c3b31 (MD5) Previous issue date: 2010-07-23 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior / The human voice is an important communication tool and any disorder of the voice can have profound implications for social and professional life of an individual. Techniques of digital signal processing have been used by acoustic analysis of vocal disorders caused by pathologies in the larynx, due to its simplicity and noninvasive nature. This work deals with the acoustic analysis of voice signals affected by pathologies in the larynx, specifically, edema, and nodules on the vocal folds. The purpose of this work is to develop a classification system of voices to help pre-diagnosis of pathologies in the larynx, as well as monitoring pharmacological treatments and after surgery. Linear Prediction Coefficients (LPC), Mel Frequency cepstral coefficients (MFCC) and the coefficients obtained through the Wavelet Packet Transform (WPT) are applied to extract relevant characteristics of the voice signal. For the classification task is used the Support Vector Machine (SVM), which aims to build optimal hyperplanes that maximize the margin of separation between the classes involved. The hyperplane generated is determined by the support vectors, which are subsets of points in these classes. According to the database used in this work, the results showed a good performance, with a hit rate of 98.46% for classification of normal and pathological voices in general, and 98.75% in the classification of diseases together: edema and nodules / A voz humana ? uma importante ferramenta de comunica??o e qualquer funcionamento inadequado da voz pode ter profundas implica??es na vida social e profissional de um indiv?duo. T?cnicas de processamento digital de sinais t?m sido utilizadas atrav?s da an?lise ac?stica de desordens vocais provocadas por patologias na laringe, devido ? sua simplicidade e natureza n?o-invasiva. Este trabalho trata da an?lise ac?stica de sinais de vozes afetadas por patologias na laringe, especificamente, edemas e n?dulos nas pregas vocais. A proposta deste trabalho ? desenvolver um sistema de classifica??o de vozes para auxiliar no pr?-diagn?stico de patologias na laringe, bem como no acompanhamento de tratamentos farmacol?gicos e p?s-cir?rgicos. Os coeficientes de Predi??o Linear (LPC), Coeficientes Cepstrais de Freq??ncia Mel (MFCC) e os coeficientes obtidos atrav?s da Transformada Wavelet Packet (WPT) s?o aplicados para extra??o de caracter?sticas relevantes do sinal de voz. ? utilizada para a tarefa de classifica??o M?quina de Vetor de Suporte (SVM), a qual tem como objetivo construir hiperplanos ?timos que maximizem a margem de separa??o entre as classes envolvidas. O hiperplano gerado ? determinado pelos vetores de suporte, que s?o subconjuntos de pontos dessas classes. De acordo com o banco de dados utilizado neste trabalho, os resultados apresentaram um bom desempenho, com taxa de acerto de 98,46% para classifica??o de vozes normais e patol?gicas em geral, e 98,75% na classifica??o de patologias entre si: edemas e n?dulos

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / The handwritten signature is a form of personal identification widely accepted, both socially and legally, used for centuries to authenticate documents such as bank checks, letters, contracts and any type of service that requires proof of authorship. When signing, an individual inserts a large amount of information to transform the movement of his hand into an identifying element. Writing speed, trajectory traversed, pen inclination, applied pressure, all these data are articulated (in the form of latent variables) to result in a static figure in the signed document. This dissertation investigates the problem of trajectory extraction of the pen that generates this signature from the resulting static image. For this, the work was divided in three main steps that are: (i) compute the skeleton of the offline signature; (ii) extract its characteristics using a concept addressed in this work called UCSS, and, with the help of previously recorded online signatures, (iii) estimate the path that the pen traveled over the skeleton. In each of these steps, a review of relevant work on the themes was done prior to begin implementation. Three experiments were carried out during this work, the first one was done with the objective of comparing the results obtained with the developed algorithm and the results obtained in a reference work, the other two were realized during the production of an article destined to the publication attached to this job. / A assinatura manuscrita é uma forma de identificação pessoal amplamente aceita, tanto social como juridicamente, utilizada há séculos para autenticar documentos como cheques bancários, cartas, contratos e todo tipo de serviço que necessite prova de autoria. Ao assinar, um indivíduo insere uma grande quantidade de informação para transformar o movimento de sua mão em um elemento identificador. Velocidade de escrita, trajetória percorrida, inclinação da caneta, pressão aplicada, todos esses dados são articulados (na forma de variáveis latentes) para resultar em uma figura estática no documento assinado. Essa dissertação investiga o problema de extração de trajetória da caneta que gera essa assinatura a partir da imagem estática resultante. Para isso, o trabalho foi divido em três passos principais que são: (i) computar o esqueleto da assinatura offline; (ii) extrair suas características utilizando um conceito abordado neste trabalho chamado UCSS, e, com o auxílio de assinaturas online registradas previamente; (iii) estimar o caminho que a caneta percorreu sobre o esqueleto. Em cada um desses passos, foi feita uma revisão de trabalhos relevantes sobre os temas para só então iniciar as implementações. Foram realizados três experimentos durante este trabalho, o primeiro foi feito com o objetivo de comparar os resultados obtidos com o algoritmo desenvolvido e os resultados obtidos em um trabalho de referência, os outros dois foram realizados durante a produção de um artigo destinado a publicação anexado a este trabalho.

Ensemble Stream Modeling and Data-cleaning are sensor information processing systems have different training and testing methods by which their goals are cross-validated. This research examines a mechanism, which seeks to extract novel patterns by generating ensembles from data. The main goal of label-less stream processing is to process the sensed events to eliminate the noises that are uncorrelated, and choose the most likely model without over fitting thus obtaining higher model confidence. Higher quality streams can be realized by combining many short streams into an ensemble which has the desired quality. The framework for the investigation is an existing data mining tool. First, to accommodate feature extraction such as a bush or natural forest-fire event we make an assumption of the burnt area (BA*), sensed ground truth as our target variable obtained from logs. Even though this is an obvious model choice the results are disappointing. The reasons for this are two: One, the histogram of fire activity is highly skewed. Two, the measured sensor parameters are highly correlated. Since using non descriptive features does not yield good results, we resort to temporal features. By doing so we carefully eliminate the averaging effects; the resulting histogram is more satisfactory and conceptual knowledge is learned from sensor streams. Second is the process of feature induction by cross-validating attributes with single or multi-target variables to minimize training error. We use F-measure score, which combines precision and accuracy to determine the false alarm rate of fire events. The multi-target data-cleaning trees use information purity of the target leaf-nodes to learn higher order features. A sensitive variance measure such as f-test is performed during each node’s split to select the best attribute. Ensemble stream model approach proved to improve when using complicated features with a simpler tree classifier. The ensemble framework for data-cleaning and the enhancements to quantify quality of fitness (30% spatial, 10% temporal, and 90% mobility reduction) of sensor led to the formation of streams for sensor-enabled applications. Which further motivates the novelty of stream quality labeling and its importance in solving vast amounts of real-time mobile streams generated today.

A recuperação de imagens baseada em conteúdo (Content-based Image Retrieval - CBIR) embasa-se sobre dois aspectos primordiais, um extrator de características o qual deve prover as características intrínsecas mais significativas dos dados e uma função de distância a qual quantifica a similaridade entre tais dados. O grande desafio é justamente como alcançar a melhor integração entre estes dois aspectos chaves com intuito de obter maior precisão nas consultas por similaridade. Apesar de inúmeros esforços serem continuamente despendidos para o desenvolvimento de novas técnicas de extração de características, muito pouca atenção tem sido direcionada à importância de uma adequada associação entre a função de distância e os extratores de características. A presente Dissertação de Mestrado foi concebida com o intuito de preencher esta lacuna. Para tal, foi realizada a análise do comportamento de diferentes funções de distância com relação a tipos distintos de vetores de características. Os três principais tipos de características intrínsecas às imagens foram analisados, com respeito a distribuição de cores, textura e forma. Além disso, foram propostas duas novas técnicas para realização de seleção de características com o desígnio de obter melhorias em relação à precisão das consultas por similaridade. A primeira técnica emprega regras de associação estatísticas e alcançou um ganho de até 38% na precisão, enquanto que a segunda técnica utilizando a entropia de Shannon alcançou um ganho de aproximadamente 71% ao mesmo tempo em que reduz significantemente a dimensionalidade dos vetores de características. O presente trabalho também demonstra que uma adequada utilização das funções de distância melhora efetivamente os resultados das consultas por similaridade. Conseqüentemente, desdobra novos caminhos para realçar a concepção de sistemas CBIR / The retrieval of images by visual content relies on a feature extractor to provide the most meaningful intrinsic characteristics (features) from the data, and a distance function to quantify the similarity between them. A challenge in this field supporting content-based image retrieval (CBIR) to answer similarity queries is how to best integrate these two key aspects. There are plenty of researching on algorithms for feature extraction of images. However, little attention have been paid to the importance of the use of a well-suited distance function associated to a feature extractor. This Master Dissertation was conceived to fill in this gap. Therefore, herein it was investigated the behavior of different distance functions regarding distinct feature vector types. The three main types of image features were evaluated, regarding color distribution, texture and shape. It was also proposed two new techniques to perform feature selection over the feature vectors, in order to improve the precision when answering similarity queries. The first technique employed statistical association rules and achieve up to 38% gain in precision, while the second one employing the Shannon entropy achieved 71%, while siginificantly reducing the size of the feature vector. This work also showed that the proper use of a distance function effectively improves the similarity query results. Therefore, it opens new ways to enhance the acceptance of CBIR systems

A taxonomia vegetal atualmente exige um grande esforço dos botânicos, desde o processo de aquisição do espécime até a morosa comparação com as amostras já catalogadas em um herbário. Nesse contexto, o projeto TreeVis surge como uma ferramenta para a identificação de vegetais por meio da análise de atributos foliares. Este trabalho é uma ramificação do projeto TreeVis e tem o objetivo de identificar vegetais por meio da análise do corte transversal de uma folha ampliado por um microscópio. Para tanto, foram extraídas assinaturas da cutícula, epiderme superior, parênquima paliçádico e parênquima lacunoso. Cada assinatura foi avaliada isoladamente por uma rede neural pelo método leave-one-out para verificar a sua capacidade de discriminar as amostras. Uma vez selecionados os vetores de características mais importantes, os mesmos foram combinados de duas maneiras. A primeira abordagem foi a simples concatenação dos vetores selecionados; a segunda, mais elaborada, reduziu a dimensionalidade (três atributos apenas) de algumas das assinaturas componentes antes de fazer a concatenação. Os vetores finais obtidos pelas duas abordagens foram testados com rede neural via leave-one-out para medir a taxa de acertos alcançada pelo sinergismo das assinaturas das diferentes partes da folha. Os experimentos consitiram na identificação de oito espécies diferentes e na identificação da espécie Gochnatia polymorpha nos ambientes Cerrado e Mata Ciliar, nas estações Chuvosa e Seca, e sob condições de Sol e Sombra / Currently, taxonomy demands a great effort from the botanists, ranging from the process of acquisition of the sample to the comparison with the species already classified in the herbarium. For this reason, the TreeVis is a project created to identify vegetal species using leaf attributes. This work is a part of the TreeVis project and aims at identifying vegetal species by analysing cross-sections of leaves amplified by a microscope. Signatures were extract from cuticle, adaxial epiderm, palisade parenchyma and sponge parenchyma. Each signature was analysed by a neural network with the leave-one-out method to verify its ability to identify species. Once the most important feature vectors were selected, two different approachs were adopted. The first was a simple concatenation of the selected feature vectors. The second, and more elaborated approach, consisted of reducing the dimensionality (three attributes only) of some component signatures before the feature vector concatenation. The final vectors obtained by these two approaches were tested by a neural network with leave-one-out to measure the correctness rate reached by the synergism of the signatures of different leaf regions. The experiments resulted in the identification of eight different species and the identification of the Gochnatia polymorpha species in Cerradão and Gallery Forest environments, Wet and Dry seasons, and under Sun and Shadow constraints

As the popularity of the master’s in data science at Dalarna University increases, so does the number of applicants. The aim of this thesis was to explore different approaches to provide an automated motivation letter scoring system which could emulate the human judgement and automate the process of candidate selection. Several steps such as image processing and text processing were required to enable the authors to retrieve numerous features which could lead to the identification of the factors graded by the program managers. Grammatical based features and Advanced textual features were extracted from the motivation letters followed by the application of Topic Modelling methods to extract the probability of each topics occurring within a motivation letter. Furthermore, correlation analysis was applied to quantify the association between the features and the different factors graded by the program managers, followed by Ordinal Logistic Regression and Random Forest to build models with the most impactful variables. Finally, Naïve Bayes Algorithm, Random Forest and Support Vector Machine were used, first for classification and then for prediction purposes. These results were not promising as the factors were not accurately identified. Nevertheless, the authors suspected that the factors may be strongly related to the highlight of specific topics within a motivation letter which can lead to further research.

Visible and near-infrared diffuse reflectance spectroscopy has been demonstrated to be a fast and cheap tool for estimating a large number of chemical and physical soil properties, and effective features extracted from spectra are crucial to correlating with these properties. We adopt a novel methodology for feature extraction of soil spectroscopy based on fractal geometry. The spectrum can be divided into multiple segments with different step–window pairs. For each segmented spectral curve, the fractal dimension value was calculated using variation estimators with power indices 0.5, 1.0 and 2.0. Thus, the fractal feature can be generated by multiplying the fractal dimension value with spectral energy. To assess and compare the performance of new generated features, we took advantage of organic soil samples from the large-scale European Land Use/Land Cover Area Frame Survey (LUCAS). Gradient-boosting regression models built using XGBoost library with soil spectral library were developed to estimate N, pH and soil organic carbon (SOC) contents. Features generated by a variogram estimator performed better than two other estimators and the principal component analysis (PCA). The estimation results for SOC were coefficient of determination (R2) = 0.85, root mean square error (RMSE) = 56.7 g/kg, the ratio of percent deviation (RPD) = 2.59; for pH: R2 = 0.82, RMSE = 0.49 g/kg, RPD = 2.31; and for N: R2 = 0.77, RMSE = 3.01 g/kg, RPD = 2.09. Even better results could be achieved when fractal features were combined with PCA components. Fractal features generated by the proposed method can improve estimation accuracies of soil properties and simultaneously maintain the original spectral curve shape.

The aim of this thesis is to explore clustering algorithms of machine unsupervised learning, which can be used for image database classification by similarity. For chosen clustering algorithms is written up a theoretical basis. For better classification of used database this thesis deals with different methods of image preprocessing. With these methods the features from image are extracted. Next the thesis solves of implementation of preprocessing methods and practical application of clustering algorithms. In practical part is programmed aplication in Python programming language, which classifies the database of images into classes by similarity. The thesis tests all of used methods and at the end of the thesis is processed searches of results.

Different music performances of the same piece may significantly differ from each other. Not only the composer and the score defines the listener’s music experience, but the music performance itself is an integral part of this experience. Four parameter classes can be used to describe a performance objectively: tempo and timing, loudness (dynamics), timbre, and pitch. Each of the individual parameters or their combination can generate a unique characteristic performance. The extraction of such objective parameters is one of the difficulties in the field of Music Performance Analysis and Music Information Retrieval. The submitted work summarizes knowledge and methods from both of the fields. The system is applied to extract data from 31 string quartet performances of 2. movement Lento of String Quartet no. 12 F major (1893) by czech romantic composer Antonín Dvořák (1841–1904).

Les micro-expressions sont des expressions faciales brèves et subtiles qui apparaissent et disparaissent en une fraction de seconde. Ce type d'expressions reflèterait "l'intention réelle" de l'être humain. Elles ont été étudiées pour mieux comprendre les communications non verbales et dans un contexte médicale lorsqu'il devient presque impossible d'engager une conversation ou d'essayer de traduire les émotions du visage ou le langage corporel d'un patient. Cependant, détecter et reconnaître les micro-expressions est une tâche difficile pour l'homme. Il peut donc être pertinent de développer des systèmes d'aide à la communication exploitant les micro-expressions. De nombreux travaux ont été réalisés dans les domaines de l'informatique affective et de la vision par ordinateur pour analyser les micro-expressions, mais une grande majorité de ces méthodes repose essentiellement sur des méthodes de vision par ordinateur classiques telles que les motifs binaires locaux, les histogrammes de gradients orientés et le flux optique. Étant donné que ce domaine de recherche est relativement nouveau, d'autres pistes restent à explorer. Dans cette thèse, nous présentons une nouvelle méthodologie pour l'analyse des petits mouvements (que nous appellerons par la suite mouvements subtils) et des micro-expressions. Nous proposons d'utiliser la pyramide de Riesz, une approximation multi-échelle et directionnelle de la transformation de Riesz qui a été utilisée pour l'amplification du mouvement dans les vidéos à l'aide de l'estimation de la phase 2D locale. Pour l'étape générale d'analyse de mouvements subtils, nous transformons une séquence d'images avec la pyramide de Riesz, extrayons et filtrons les variations de phase de l'image. Ces variations de phase sont en lien avec le mouvement. De plus, nous isolons les régions d'intérêt où des mouvements subtils pourraient avoir lieu en masquant les zones de bruit à l'aide de l'amplitude locale. La séquence d'image est transformée en un signal ID utilisé pour l'analyse temporelle et la détection de mouvement subtils. Nous avons créé notre propre base de données de séquences de mouvements subtils pour tester notre méthode. Pour l'étape de détection de micro-expressions, nous adaptons la méthode précédente au traitement de certaines régions d'intérêt du visage. Nous développons également une méthode heuristique pour détecter les micro-événements faciaux qui sépare les micro-expressions réelles des clignotements et des mouvements subtils des yeux. Pour la classification des micro-expressions, nous exploitons l'invariance, sur de courtes durées, de l'orientation dominante issue de la transformation de Riesz afin de moyenner la séquence d'une micro-expression en une paire d'images. A partir de ces images, nous définissons le descripteur MORF (Mean Oriented Riesz Feature) constitué d'histogrammes d'orientation. Les performances de nos méthodes sont évaluées à l'aide de deux bases de données de micro-expressions spontanées. / Micro-expressions are brief and subtle facial expressions that go on and off the face in a fraction of a second. This kind of facial expressions usually occurs in high stake situations and is considered to reflect a humans real intent. They have been studied to better understand non-verbal communications and in medical applications where is almost impossible to engage in a conversation or try to read the facial emotions or body language of a patient. There has been some interest works in micro-expression analysis, however, a great majority of these methods are based on classically established computer vision methods such as local binary patterns, histogram of gradients and optical flow. Considering the fact that this area of research is relatively new, much contributions remains to be made. ln this thesis, we present a novel methodology for subtle motion and micro-expression analysis. We propose to use the Riesz pyramid, a multi-scale steerable Hilbert transformer which has been used for 2-D phase representation and video amplification, as the basis for our methodology. For the general subtle motion analysis step, we transform an image sequence with the Riesz pyramid, extract and lifter the image phase variations as proxies for motion. Furthermore, we isolate regions of intcrcst where subtle motion might take place and mask noisy areas by thresholding the local amplitude. The total sequence is transformed into a ID signal which is used fo temporal analysis and subtle motion spotting. We create our own database of subtle motion sequences to test our method. For the micro-expression spotting step, we adapt the previous method to process some facial regions of interest. We also develop a heuristic method to detect facial micro-events that separates real micro-expressions from eye blinkings and subtle eye movements. For the micro-expression classification step, we exploit the dominant orientation constancy fom the Riesz transform to average the micro-expression sequence into an image pair. Based on that, we introduce the Mean Oriented Riesz Feature descriptor. The accuracy of our methods are tested in Iwo spontaneous micro-expressions databases. Furthermore, wc analyse the parameter variations and their effect in our results.