L'analyse automatique de contenu des vidéos en vue de leur annotation est un domaine de recherche en plein essor. Reconnaître les personnes apparaissant dans des émissions audiovisuelles permet une structuration automatique d'une quantité grandissante d'archives audiovisuelles. Nous présentons une approche dynamique originale de reconnaissance de personnes dans les flux vidéo. Cette approche est dynamique car elle tire avantage de la richesse des informations contenues dans la vidéo, contrairement aux approches statiques basées uniquement sur les images. L'approche proposée comprend deux volets. Le premier volet consiste à isoler toutes les occurrences de personnes d'une émission, et à les regrouper en clusters en se basant sur un descripteur original : les histogrammes spatio-temporels, ainsi que sur une mesure de similarité dédiée. L'originalité vient de l'intégration d'informations temporelles dans le descripteur, qui permet une estimation plus fiable de la similarité entre les occurrences de personnes. Le second volet propose la mise en oeuvre d'une méthode de reconnaissance faciale. Différentes stratégies sont envisagées, d'une part pour identifier les occurrences de personnes selon les trames qui composent la séquence, et d'autre part pour propager les identités au sein des groupes selon leurs membres. Ces deux aspects de notre contribution ont été évalués à l'aide de corpus de données réelles contenant des émissions issues des chaînes BFMTV et LCP. Les résultats des expérimentations menées indiquent que l'approche proposée permet d'améliorer notablement la précision de reconnaissance en prenant en compte la dimension temporelle. / The annotation of video streams by automatic content analysis is a growing field of research. The possibility of recognising persons appearing in TV shows allows to automatically structure ever-growing video archives. We present an original and dynamic approach to person recognition from video streams. This approach is dynamic as it benefits from the motion information contained in videos, whereas the static approaches are solely based on still images. The proposed approach is composed of two parts. In the first one, we extract persontracks from the shows and cluster them using a new descriptor and its associated similarity measure : space-time histograms. The originality of our approach is the integration of temporal data into the descriptor. Experiments show that it provides a better estimation of the similarity between persontracks. In the second part of our approach, we propose to use a facial recognition method which aims at "naming" the clusters. Various strategies are considered to assign an identity to a persontrack using its frames and to propagate this identity to members of the same cluster. These two aspects of our contribution have been evaluated using a corpus of real life TV shows broadcasted on BFMTV and LCP TV channels. The results of our experiments show that our approach significantly improves the precision of the recognition process thanks to the use of the temporal dimension.

This thesis concentrates on learning and identification of fuzzy systems, and this thesis is composed about learning fuzzy systems from data for regression and function approximation by constructing complete, compact, and consistent fuzzy systems. Fuzzy systems are prevalent to solve pattern recognition problems and function approximation problems as a result of the good knowledge representation. With the development of fuzzy systems, a lot of sophisticated methods based on them try to completely solve pattern recognition problems and function approximation problems by constructing a great diversity of mathematical models. However, there exists a conflict between the degree of the interpretability and the accuracy of the approximation in general fuzzy systems. Thus, how to properly make the best compromise between the accuracy of the approximation and the degree of the interpretability in the entire system is a significant study of the subject.The first work of this research is concerned with the clustering technique on constructing fuzzy models in fuzzy system identification, and this method is a part of clustering based learning of fuzzy systems. As the determination of the proper number of clusters and the appropriate location of clusters is one of primary considerations on constructing an effectively fuzzy model, the task of the clustering technique aims at recognizing the proper number of clusters and the appropriate location as far as possible, which gives a good preparation for the construction of fuzzy models. In order to acquire the mutually exclusive performance by constructing effectively fuzzy models, a modular method to fuzzy system identification based on a hybrid clustering-based technique has been considered. Due to the above reasons, a hybrid clustering algorithm concerning input, output, generalization and specialization has hence been introduced in this work. Thus, the primary advantage of this work is the proposed clustering technique integrates a variety of clustering properties to positively identify the proper number of clusters and the appropriate location of clusters by carrying out a good performance of recognizing the precise position of each dataset, and this advantage brings fuzzy systems more complete.The second work of this research is an extended work of the first work, and two ways to improve the original work have been considered in the extended work, including the pruning strategy for simplifying the structure of fuzzy systems and the optimization scheme for parameters optimization. So far as the pruning strategy is concerned, the purpose of which aims at refining rule base by the similarity analysis of fuzzy sets, fuzzy numbers, fuzzy membership functions or fuzzy rules. By other means, through the similarity analysis of which, the complete rules can be kept and the redundant rules can be reduced probably in the rule base of fuzzy systems. Also, the optimization scheme can be regarded as a two-layer parameters optimization in the extended work, because the parameters of the initial fuzzy model have been fine tuning by two phases gradation on layer. Hence, the extended work primarily puts focus on enhancing the performance of the initial fuzzy models toward the positive reliability of the final fuzzy models. Thus, the primary advantage of this work consists of the simplification of fuzzy rule base by the similarity-based pruning strategy, as well as more accuracy of the optimization by the two-layer optimization scheme, and these advantages bring fuzzy systems more compact and precise.So far as a perfect modular method for fuzzy system identification is concerned, in addition to positively solve pattern recognition problems and function approximation problems, it should primarily comprise the following features, including the well-understanding interpretability, low-degree dimensionality, highly reliability, stable robustness, highly accuracy of the approximation, less computational cost, and maximum performance. However, it is extremely difficult to meet all of these conditions above. Inasmuch as attaining the highly achievement from the features above as far as possible, the research works of this thesis try to present a modular method concerning a variety of requirements to fuzzy systems identification.

This study examined problems referred to Teacher Assistance Teams (TATs) during the 1988-89 school year in four Vancouver schools. Exploratory analyses were conducted to discover: (a) similarities or differences between problems referred to Teacher Assistance Teams and those referred to School-based teams (SBTs); and (b) similarities or differences between problem statements before and after the problem identification phase of the TAT process was carried out. Significant differences were found in the nature of problems referred to TATs and to SBTs. No significant differences were found in problem statements before and after problem identification was carried out. / Education, Faculty of / Graduate

Ph.D. (Computer Science) / The introduction of new portable sensors that monitor physiological systems in the human body has allowed quality of life and medical diagnostic applications to be taken directly to the user, without the constraints of physical space or inconvenience. The potential of these sensors in the domain of authentication and identi cation is becoming more feasible each day and current research in these biometric systems show a great deal of promise. Novel biometric systems are being introduced that use biological signals (also known as biosignals) in the human body captured by these sensors (such as brain waves or heart rate) as the core unique attribute. The study builds on the proliferation of these sensors and proposes an interoperable model called CoBI, which allows individual or multi-factor authentication and identi cation to take place. The model provides a platform for any viable biosignal that can be used for the purposes of identi cation and authentication, by providing pluggable sensor and signal processing components. These components can then convert biosignals into a common format, a feature vector consisting of estimated autoregressive (AR) coe cients. Once they are in a common format they can then be merged together to form a consolidated feature vector using feature fusion. This consolidated feature vector can then be persisted during enrolment or passed further for matching using classi cation techniques, such as K-Nearest Neighbour. The results, from the comprehensive benchmark performed (called BAMBI) on an implemented version of the model (called CaNViS), have shown that biological signals that contain cardiac and neurological components (ie. from an electrocardiogram (ECG) and electroencephalogram (EEG), respectively) can be captured, processed, consolidated and classi ed using the CoBI model successfully. By utilising the correct AR model order during feature estimation for the cardiac and neurological components, along with the appropriate classi er for matching, the biometric system yields nominal results for authentication and identi cation in access control environments.

This investigation includes a morphological and cultural study of four rather common pathogenic fungi one of which attacks Zinnia elegans, one Lagerstroemia indica, one Ligustrum lodense and one Euonymus japonica. The problem includes, first, a determination of the genera of these four fungi as revealed by their morphology in natural habitat; second, a determination of the growth and cultural characteristics of these on Cornmeal, Potato, Bean, Prune and Carrot agars; third, a comparison of the growth and cultural characteristics of these in natural habitat with the same on the various agars used; and fourth, an evaluation of the five artificial culture media in so far as their usage is concerned in the growth and cultivation of these four species of fungi.

The purpose of the present study was to determine the relationship between masculine traits in males and the characteristic patterns of emotional responses which affect social adjustment.

Orientadores: Eduardo Daruge, Eduardo Daruge Júnior / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Odontologia de Piracicaba / Made available in DSpace on 2018-08-18T13:57:11Z (GMT). No. of bitstreams: 1 Martins_EmersonClaudino_M.pdf: 705232 bytes, checksum: cd0617008d8a27566e3ef958bb603094 (MD5) Previous issue date: 2011 / Resumo: No presente trabalho estudou-se 139 imagens tomográficas, obtidas dos arquivos do Hospital das Clínicas de Rubião Júnior da Faculdade de Medicina de Botucatu - UNESP, sendo 76 do gênero masculino e 63 do gênero feminino. Foram realizadas medidas quantitativas das imagens, nos cortes transversais entre dois pontos determinados nas extremidades limites radiopacas do Manúbrio e do Processo Xifóide, por meio de tomografias computadorizadas em adultos com idades variando de 30 a 95 anos. Verificou-se que para o Processo Xifóide houve diferenças entre a faixa etária e o gênero, com um aumento quantitativo nas medidas, mas para o manúbrio houve diferença apenas entre o gênero na faixa etária de 50 a 95 anos e não deu diferença entre as faixas etárias. Conclui-se que o método de estimativa da idade em indivíduos adultos mensurando-se a imagem de osso esterno nas sinostoses por meio da tomografia computadorizada é uma alternativa viável para se indicar a idade cronológica de um indivíduo com 50 anos ou mais / Abstract: In the present work it was studied 139 images tomográficas, obtained of the files of the Hospital of the Clinics of Rubião Júnior of University of Medicine of Botucatu - UNESP, being 76 of the masculine gender and 63 of the feminine gender. Quantitative measures of the images were accomplished, in the traverse cuts among two certain points in the extremities limits radiopacas of Manúbrio and of the Processo Xifóide, through computerized tomographies in adults with ages varying from 30 to 95 years. It was verified that there were differences between the age group and the gender for the Processo Xifóide, with a quantitative increase in the measures, but to stop the manúbrio there was differentiates just among the gender in the age group from 50 to 95 years and he/she didn't give difference among the age groups. It is ended that the method of estimate of the age in adult individuals being measured the image of bone breastbone in the sinostoses through the computerized tomography is a viable alternative to indicate an individual's chronological age with 50 years or more / Mestrado / Odontologia Legal e Deontologia / Mestre em Biologia Buco-Dental

Orientador: Plínio Trabasso / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Ciências Médicas / Made available in DSpace on 2018-08-24T11:49:19Z (GMT). No. of bitstreams: 1 Souza_Marcelade_M.pdf: 1154397 bytes, checksum: c435b972ac36d284080e07e6b6b20a6e (MD5) Previous issue date: 2014 / Resumo: Este trabalho teve por objetivo comparar o desempenho das técnicas de microarranjo de DNA (DNA microarray), amplificação circular isotérmica (Loop mediated isothermal amplification (LAMP)), PCR em tempo real e sequenciamento de DNA para a identificação das espécies de Fusarium, agentes de infecção da corrente sanguínea, isolados de pacientes com doenças onco-hematológicas do Hospital de Clínicas da Universidade Estadual de Campinas (HC-Unicamp) e do Hospital de Clínicas da Faculdade de Medicina da Universidade de São Paulo (FMUSP) em um estudo de corte retrospectivo. Foram selecionados 21 amostras de Fusarium isoladas de sangue, sendo 17 amostras da micoteca do Laboratório de Investigação em Fungos (LIF) da Faculdade de Ciências Médicas da Unicamp (FCM-Unicamp) e três amostras da micoteca do Laboratório de Microbiologia da FMUSP. As metodologias de DNA microarray, LAMP e PCR em tempo real, identificaram 17 das 21 amostras como complexo Fusarium solani e quatro como complexo Fusarium não - solani, apresentando um alto desempenho, uma concordância de 100% entre elas e uma sensibilidade de 100% para as metodologias DNA microarray, LAMP e PCR em tempo real e uma especificidade de 100% para a metodologia de LAMP. Esses dados foram validados pela metodologia de sequenciamento de DNA, a qual identificou 17 amostras como complexo Fusarium solani e quatro como Fusarium complexo não ¿ solani, sendo três Fusarium napiforme e um Fusarium oxysporum, concordando com os resultados encontrados nas três metodologias aplicadas neste trabalho. A técnica de LAMP demonstrou ser mais acessível do que as técnicas de DNA microarray e PCR em tempo real por ser mais rápida, indicando ser uma metodologia promissora para ser utilizada na rotina de diagnóstico de fungos filamentosos, auxiliando em uma terapia apropriada de uma forma rápida e específica, proporcionando assim uma melhora na assistência prestada ao paciente / Abstract: This study aimed to compare the performance of the techniques of DNA microarray, loop mediated isothermal amplification (LAMP), real-time PCR and DNA sequencing for identification of species of Fusarium as agents of bloodstream infection, isolated from patients with hematologic malignancies cared for at the university hospital of State University of Campinas and the university hospital of University of Sao Paulo, in a retrospective cross study. In the study, 21 Fusarium isolates from blood were selected, being 17 samples of mycology collection of the Laboratory for Research on Fungi (LIF), Faculty of Medical Sciences, Campinas (FCM-Unicamp) and three samples of mycology collection of the Laboratory of Microbiology of FMUSP. The techniques of DNA microarray, LAMP and Real-Time PCR identified 17 of the 21 samples as Fusarium solani, and 4 isolates as Fusarium non-solani complex, showing a high performance, a 100% agreement among them and a sensitivity of 100% for the DNA microarray, LAMP and real-time PCR methods and a 100% specificity for the LAMP method. These data were validated by the DNA sequencing technique, which identified 17 isolates as Fusarium solani complex and not four as Fusarium non-solani complex, being three Fusarium napiforme and one Fusarium oxysporum, agreeing with the findings in the three methodologies applied in this work. The LAMP technique proved to be more accessible than the techniques of DNA microarray and real-time PCR to be faster, indicating a promising methodology to be used in routine diagnosis of filamentous fungi, aiding in proper therapy rapid and specific manner, thereby providing an improvement to the care of patients / Mestrado / Clinica Medica / Mestra em Clínica Médica

Misleading post-event information--Cognitive processing differences.

The objectives of this study are to identify the larval forms of the two species of smelt, Spirinchus thaleichthys and Hypomesus transpacificus transpacificus, and to better describe their spawning times in the Sacramento-San Joaquin Estuary.