Investigation into DCT Feature Selection for Visual Lip-Based Biometric Authentication

This paper investigated using lip movements as a behavioural biometric for person authentication. The system was trained, evaluated and tested using the XM2VTS dataset, following the Lausanne Protocol configuration II. Features were selected from the DCT coefficients of the greyscale lip image. This paper investigated the number of DCT coefficients selected, the selection process, and static and dynamic feature combinations. Using a Gaussian Mixture Model - Universal Background Model framework an Equal Error Rate of 2.20% was achieved during evaluation and on an unseen test set a False Acceptance Rate of 1.7% and False Rejection Rate of 3.0% was achieved. This compares favourably with face authentication results on the same dataset whilst not being susceptible to spoofing attacks.

k-fold Subsampling based Sequential Backward Feature Elimination

We present a new wrapper feature selection algorithm for human detection. This algorithm is a hybrid featureselection approach combining the benefits of filter and wrapper methods. It allows the selection of an optimalfeature vector that well represents the shapes of the subjects in the images. In detail, the proposed featureselection algorithm adopts the k-fold subsampling and sequential backward elimination approach, while thestandard linear support vector machine (SVM) is used as the classifier for human detection. We apply theproposed algorithm to the publicly accessible INRIA and ETH pedestrian full image datasets with the PASCALVOC evaluation criteria. Compared to other state of the arts algorithms, our feature selection based approachcan improve the detection speed of the SVM classifier by over 50% with up to 2% better detection accuracy.Our algorithm also outperforms the equivalent systems introduced in the deformable part model approach witharound 9% improvement in the detection accuracy

Retratos de família: a identidade nacional e a violência em João Canijo

Esta dissertação pretende analisar a obra de João Canijo e a sua relação com as representações da identidade nacional. Para isso, socorre-se de uma revisão bibliográfica sobre a identidade cultural portuguesa, em diversas dimensões (histórica, literária e antropológica), ressaltando, sobretudo, a importância da ideologia salazarista e a tensão identitária da situação atual. Num segundo momento, este trabalho preocupa-se com a história do cinema português, desde os anos 60, procurando estabelecer paradigmas e mudanças, sobretudo na relação com o imaginário português. Na terceira parte, a dissertação ensaia uma análise a oito longas-metragens do realizador, propondo a ideia de uma dramaturgia da violência, através de uma intertextualidade com a tragédia grega e o melodrama cinematográfico, que pretende dar conta de um imaginário português contemporâneo, que se baseia numa sociedade patriarcal e na sua violência. Nesse sentido, argumenta-se a importância de conceitos como a “não-inscrição”, de José Gil, ou o recalcado, de Eduardo Lourenço, para a compreensão dos filmes. Num último momento, esta análise percorre o debate do realismo no cinema, através do prisma das mudanças contemporâneas sugeridas pela obra do cineasta, em que se destaca uma hibridez entre elementos ficcionais e documentais. Sugere-se, finalmente, uma dramaturgia da violência nos filmes de João Canijo que procura rever o imaginário existente no que diz respeito à identidade portuguesa e às suas representações culturais.

Classification of incomplete feature vectors by radial basis function networks

The paper describes the use of radial basis function neural networks with Gaussian basis functions to classify incomplete feature vectors. The method uses the fact that any marginal distribution of a Gaussian distribution can be determined from the mean vector and covariance matrix of the joint distribution.

Feature Engineering in Fine-Grained Image Classification

Thesis (Ph.D.)--University of Washington, 2013

As freguesias e as políticas educativas locais: as necessidades e as obrigações

Dissertação apresentada para obtenção do grau de Mestre em Ciências da Educação Área de especialização em Administração Escolar 2013

Incremental filter and wrapper approaches for feature discretization

Discrete data representations are necessary, or at least convenient, in many machine learning problems. While feature selection (FS) techniques aim at finding relevant subsets of features, the goal of feature discretization (FD) is to find concise (quantized) data representations, adequate for the learning task at hand. In this paper, we propose two incremental methods for FD. The first method belongs to the filter family, in which the quality of the discretization is assessed by a (supervised or unsupervised) relevance criterion. The second method is a wrapper, where discretized features are assessed using a classifier. Both methods can be coupled with any static (unsupervised or supervised) discretization procedure and can be used to perform FS as pre-processing or post-processing stages. The proposed methods attain efficient representations suitable for binary and multi-class problems with different types of data, being competitive with existing methods. Moreover, using well-known FS methods with the features discretized by our techniques leads to better accuracy than with the features discretized by other methods or with the original features. (C) 2013 Elsevier B.V. All rights reserved.

Where's Albania? : staking out the politics of the real and reality in documentary cinema

This thesis, entitled, "Where's Albania? Staking Out the Politics of the Real and Reality in Documentary Cinema," charts the documentary tradition's path from its first incarnations, as filmed travelogue or ethnographic study, for example, right through to its development as a form acting as an objective observer, reflexive commentator, and finally, as a postmodern hybrid. This thesis begins by locating the documentary tradition's origins in realism. Foregrounding documentary cinema as a realist style is important in that it is a contention that spans this entire study. After working through the numerous modes of documentary as outlined by Bill Nichols, I suggest the documentary is often best understood as a hybrid form drawing on numerous modes and conventions. This argument permits my study to make a shift into postmodern theory, wherein I examine postmodernism's relationship to the documentary both as being influenced by it, but also as subsequently forcing documentary cinema to look back at itself and reevaluate the claims it has made in the past, and how postmodernism has drawn these claims to the surface of debate. My thesis concludes with a study of the mockumentary. This analysis confirms the link between postmodernism and documentary, but perhaps more importantly, this analysis investigates postmodemism's critique of the image and representation in general, two elements historically linked to documentary cinema's success as "truth teller."

Evolutionary synthesis of stochastic gene network models using feature-based search spaces

A feature-based fitness function is applied in a genetic programming system to synthesize stochastic gene regulatory network models whose behaviour is defined by a time course of protein expression levels. Typically, when targeting time series data, the fitness function is based on a sum-of-errors involving the values of the fluctuating signal. While this approach is successful in many instances, its performance can deteriorate in the presence of noise. This thesis explores a fitness measure determined from a set of statistical features characterizing the time series' sequence of values, rather than the actual values themselves. Through a series of experiments involving symbolic regression with added noise and gene regulatory network models based on the stochastic 'if-calculus, it is shown to successfully target oscillating and non-oscillating signals. This practical and versatile fitness function offers an alternate approach, worthy of consideration for use in algorithms that evaluate noisy or stochastic behaviour.

Feature Selection and Classification Using Age Layered Population Structure Genetic Programming

The curse of dimensionality is a major problem in the fields of machine learning, data mining and knowledge discovery. Exhaustive search for the most optimal subset of relevant features from a high dimensional dataset is NP hard. Sub–optimal population based stochastic algorithms such as GP and GA are good choices for searching through large search spaces, and are usually more feasible than exhaustive and deterministic search algorithms. On the other hand, population based stochastic algorithms often suffer from premature convergence on mediocre sub–optimal solutions. The Age Layered Population Structure (ALPS) is a novel metaheuristic for overcoming the problem of premature convergence in evolutionary algorithms, and for improving search in the fitness landscape. The ALPS paradigm uses an age–measure to control breeding and competition between individuals in the population. This thesis uses a modification of the ALPS GP strategy called Feature Selection ALPS (FSALPS) for feature subset selection and classification of varied supervised learning tasks. FSALPS uses a novel frequency count system to rank features in the GP population based on evolved feature frequencies. The ranked features are translated into probabilities, which are used to control evolutionary processes such as terminal–symbol selection for the construction of GP trees/sub-trees. The FSALPS metaheuristic continuously refines the feature subset selection process whiles simultaneously evolving efficient classifiers through a non–converging evolutionary process that favors selection of features with high discrimination of class labels. We investigated and compared the performance of canonical GP, ALPS and FSALPS on high–dimensional benchmark classification datasets, including a hyperspectral image. Using Tukey’s HSD ANOVA test at a 95% confidence interval, ALPS and FSALPS dominated canonical GP in evolving smaller but efficient trees with less bloat expressions. FSALPS significantly outperformed canonical GP and ALPS and some reported feature selection strategies in related literature on dimensionality reduction.

Feature Selection and Classification Using Age Layered Population Structure Genetic Programming

The curse of dimensionality is a major problem in the fields of machine learning, data mining and knowledge discovery. Exhaustive search for the most optimal subset of relevant features from a high dimensional dataset is NP hard. Sub–optimal population based stochastic algorithms such as GP and GA are good choices for searching through large search spaces, and are usually more feasible than exhaustive and determinis- tic search algorithms. On the other hand, population based stochastic algorithms often suffer from premature convergence on mediocre sub–optimal solutions. The Age Layered Population Structure (ALPS) is a novel meta–heuristic for overcoming the problem of premature convergence in evolutionary algorithms, and for improving search in the fitness landscape. The ALPS paradigm uses an age–measure to control breeding and competition between individuals in the population. This thesis uses a modification of the ALPS GP strategy called Feature Selection ALPS (FSALPS) for feature subset selection and classification of varied supervised learning tasks. FSALPS uses a novel frequency count system to rank features in the GP population based on evolved feature frequencies. The ranked features are translated into probabilities, which are used to control evolutionary processes such as terminal–symbol selection for the construction of GP trees/sub-trees. The FSALPS meta–heuristic continuously refines the feature subset selection process whiles simultaneously evolving efficient classifiers through a non–converging evolutionary process that favors selection of features with high discrimination of class labels. We investigated and compared the performance of canonical GP, ALPS and FSALPS on high–dimensional benchmark classification datasets, including a hyperspectral image. Using Tukey’s HSD ANOVA test at a 95% confidence interval, ALPS and FSALPS dominated canonical GP in evolving smaller but efficient trees with less bloat expressions. FSALPS significantly outperformed canonical GP and ALPS and some reported feature selection strategies in related literature on dimensionality reduction.

New Feature at Niagara – Clark Hill Islands

New Feature at Niagara – Clark Hill Islands (5 islands situated in the rapids of the Niagara River). These islands are currently known as Dufferin Islands, 22 ½ cm. x 15 ½ cm, n.d.

The Credibility and Authoritativeness of Documentary Information in Determining Refugee Status: The Canadian Experience

This article reviews the origins of the Documentation, Information and Research Branch (the 'Documentation Center') of Canada's Immigration and Refugee Board (IRB), established in 1988 as a part of a major revision of the procedure for determination of refugee status. The Documentation Center conducts research to produce documents describing conditions in refugee-producing countries, and also disseminates information from outside. The information is available to decision-makers, IRB staff, counsel and claimants. Given the importance of decisions on refugee status, the article looks at the credibility and the authoritativeness of the information, by analyzing the structure of information used. It recalls the different types of information 'package' produced, such as a country profiles and the Question and Answer Series, the Weekly Madia Review, the 'Perspectives' series, Responses to Information Requests and Country files, and considers the trend towards standardization across the country. The research process is reviewed, as are the hiring criteria for researchers, the composition of the 'collection', how acquisitions are made, and the development of databases, particularly on country of origin (human rights material) and legal information, which are accessible on-line. The author examines how documentary information can be used by decision-makers to draw conclusions as to whether the claim has a credible basis or the claimant has a well-founded fear of persecution. Relevant caselaw is available to assess and weigh the claim. The experience of Amnesty International in similar work is cited for comparative purposes. A number of 'safeguards' are mentioned, which contribute to the goal of impartiality in research, or which otherwise enhance the credibility of the information, and the author suggests that guidelines might be drafted to explain and assist in the realization of these aims. Greater resources might also enable the Center to undertake the task of 'certifying' the authoritativeness of sources. The author concludes that, as a new institution in Canadian administrative law, the Documentation Center opens interesting avenues for the future. Beacause it ensures an acceptable degree of impartiality of its research and the documents it produces, it may be a useful model for others tribunals adjudicating in fields where evidence is either difficult to gather, or is otherwise complex.

Un monde de silence : musique pour documentaires aquatiques.

La version intégrale de cette thèse est disponible uniquement pour consultation individuelle à la Bibliothèque de musique de l’Université de Montréal (http://www.bib.umontreal.ca/MU).