Thesis (Ph.D.)--University of Wollongong, 2007. / Typescript. Includes bibliographical references: leaf 274-294.

Mémoire (M.A.)-- Université du Québec à Chicoutimi, 1988. / Ce travail de recherche a été réalisé à l'Université du Québec à Chicoutimi dans le cadre du programme de maîtrise en arts plastiques extensionné de l'Université du Québec à Montréal à l'Université du Québec à Chicoutimi. CaQCU Bibliographie : p. 29-32. Document électronique également accessible en format PDF. CaQCU

Orientador: Fernando José Von Zuben / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação / Made available in DSpace on 2018-08-16T10:52:32Z (GMT). No. of bitstreams: 1 Violato_RicardoParanhosVelloso_M.pdf: 5520106 bytes, checksum: 47f6f741b5c013a3252e50dddb37923c (MD5) Previous issue date: 2010 / Resumo: Este trabalho apresenta um estudo comparativo de três algoritmos de quantização vetorial, aplicados para a compressão de sinais de fala: k-médias, NG (do inglês Neural-Gas) e ARIA. Na técnica de compressão utilizada, os sinais são primeiramente parametrizados e quantizados, para serem armazenados e/ou transmitidos. Para recompor o sinal, os vetores quantizados são mapeados em quadros de fala, que são, por sua vez, concatenados, através de uma técnica de síntese concatenativa. Esse sistema pressupõe a existência de um dicionário (codebook) de vetores-padrão (codevectors), os quais são utilizados na etapa de codificação, e de um dicionário de quadros, que é utilizado na etapa de decodificação. Tais dicionários são gerados aplicando-se um algoritmo de quantização vetorial juntoa uma base de treinamento. Em particular, deseja-se avaliar o algoritmo imuno-inspirado denominado ARIA e sua capacidade de preservação da densidade da distribuição dos dados. São testados também diferentes conjuntos de parâmetros para identificar aquele que produz os melhores resultados. Por fim, são propostas modificações no algoritmo ARIA visando ganho de desempenho tanto na preservação de densidade quanto na qualidade do sinal sintetizado / Abstract: This work presents a comparative study of three algorithms for vector quantization, applied for the compression of speech signals: k-means, NG (Neural-Gas) and ARIA. In the compression technique used, the signals are first parameterized and quantized to be stored and/or transmitted. To reconstruct the signal, the quantized vectors are mapped into speech frames, which are concatenated through a concatenative synthesis technique. This system assumes the existence of a dictionary (codebook) of reference vectors (codevectors), which is used in the coding step, and a dictionary of frames, which is used in the decoding step. These dictionaries are generated by applying a vector quantization algorithm within a training database. In particular, we want to evaluate the immune-inspired algorithm called ARIA and its ability to preserve the density of data distribution. Different sets of parameters are also tested in order to identify the one that produces the best results. Finally, modifications to the ARIA algorithm are proposed aiming at obtaining gain in performance in both the preservation of density and the quality of the synthesized signal / Mestrado / Engenharia de Computação / Mestre em Engenharia Elétrica

The goal of this thesis was to propose an image segmentation method, which is capable of effective segmentation process with small datasets. Recently published ODE neural network was used for this method, because its features should provide better generalization in case of tasks with only small datasets available. The proposed ODE-UNet network was created by combining UNet architecture with ODE neural network, while using benefits of both networks. ODE-UNet reached following results on ISBI dataset: Rand: 0,950272 and Info: 0,978061. These results are better than the ones received from UNet model, which was also tested in this thesis, but it has been proven that state of the art can not be outperformed using ODE neural networks. However, the advantages of ODE neural network over tested UNet architecture and other methods were confirmed, and there is still a room for improvement by extending this method.

Tato práce pojednává o technikách automatické analýzy obrazu založené na konvolučních neuronových sítích (CNN), zaměřených na klasifikaci planktonu. V oblasti studování planktonu panuje velká diverzita v jeho tvarech a velikostech. Kvůli tomuto bývá klasifikace pomocí CNN náročná, jelikož CNN typicky požadují definovanou velikost vstupu. Běžné metody využívají škálování obrazu do jednotné velikosti. Avšak kvůli tomuto jsou ztraceny drobné detaily potřebné ke správné klasifikaci. Cílem práce bylo navrhnout a implementovat CNN klasifikátor obrazových dat planktonu a prozkoumat metody, které jsou zaměřené na problematiku různorodých velikostí obrázků. Metody, jako jsou patch cropping, využití spatial pyramid pooling vrstvy, zahrnutí metadat a sestavení multi-stream modelu jsou vyhodnoceny na náročném datasetu obrázků fytoplanktonu. Takto bylo dosaženo zlepšení o 1.0 bodů pro InceptionV3 architekturu s výslednou úspěšností 96.2 %. Hlavním přínosem této práce je vylepšení CNN klasifikátorů planktonu díky úspěšné aplikaci těchto metod.

Multidimenzionální indexování je účinným nástrojem pro zachycení podobností mezi objekty bez nutnosti jejich explicitní kategorizace. V posledních letech byla tato metoda hojně využívána pro anotaci objektů a tvořila významnou část publikací spojených s projektem Visipedia. Tato práce analyzuje možnosti strojového učení z multidimenzionálně indexovaných obrázků na základě jejich obrazových příznaků a přestavuje metody predikce multidimenzionálních souřadnic pro předem neznámé obrázky. Práce studuje příslušené algoritmy pro extrakci příznaků, analyzuje relevantní metody strojového účení a popisuje celý proces vývoje takového systému. Výsledný systém je pak otestován na dvou různých datasetech a provedené experimenty prezentují první výsledky pro úlohu svého druhu.

Breast cancer is one of the most common cancers among women in the world, and the average error rate among radiologists during diagnosis is 30%. Computer-aided medical diagnosis aims to assist doctors by giving them a second opinion, thus decreasing the error rate. Convolutional neural networks (CNNs) have shown to be good for visual detection and recognition tasks, and have been explored in combination with transfer learning. However, the performance of a deep learning model does not only rely on the model itself, but on the nature of the dataset as well In breast cancer diagnosis, the area surrounding a mass provides useful context for diagnosis. In this study, we explore providing different amounts of context to the CNN model ResNet50, to see how it affects the model’s performance. We test masses with no additional context, twice the amount of original context and four times the amount of original context, using 10-fold cross-validation with ROC AUC and average precision (AP ) as our metrics. The results suggest that providing additional context does improve the model’s performance. However, giving two and four times the amount of context seems to give similar performance. / Bröstcancer är en av de vanligaste cancersjukdomar bland kvinnor i världen, och den genomsnittliga felfrekvensen under diagnoser är 30%. Datorstödd medicinsk diagnos syftar till att hjälpa läkare genom att ge dem en andra åsikt, vilket minskar felfrekvensen. Konvolutionella neurala nätverk (CNNs) har visat sig vara bra för visuell detektering och igenkännande, och har utforskats i samband med det s.k. “transfer learning”. Prestationen av en djup inlärningsmodell är däremot inte enbart beroende på modellen utan också på datasetets natur. I bröstcancerdiagnos ger området runt en bröstmassa användbar kontext för diagnos. I den här studien testar vi att ge olika mängder kontext till CNNmodellen ResNet50, för att se hur det påverkar modellens prestanda. Vi testar bröstmassor utan ytterligare kontext, dubbelt så mycket som den originala mängden kontext och fyra gånger så mycket som den orginala mängden kontext, med hjälp av “10-fold cross-validation” med ROC AUC och “average precision” (AP ) som våra mätvärden. Resultaten visar att mer kontext förbättrar modellens prestanda. Däremot verkar att ge två och fyra gånger så mycket kontext resultera i liknande prestanda.

Made available in DSpace on 2016-04-28T18:24:07Z (GMT). No. of bitstreams: 1 Luciana Costa Nogueira.pdf: 903522 bytes, checksum: be9ecf4748c39381268323ea4f8bff02 (MD5) Previous issue date: 2009-07-28 / Secretaria da Educação do Estado de São Paulo / The objectives of this study are: to analyze how the teaching-learning theories and language descriptions that ground the Brazilian National Curriculum Parameters to teach foreign languages in elementary and secondary schools (PCN-LE, Brasil, 1998 and PCN-EM, Brasil, 1999) are present in a computer-mediated English course called Interaction Teachers; and to identify the perceptions of teachers, participants of this course, on linguistic and technological aspects exploited in it. This research is grounded on concepts discussed in the Brazilian National Curriculum Parameters to teach foreign languages in elementary and secondary schools (PCN-LE, Brasil, 1998 and PCN-EM, Brazil, 1999) as well as on currents of thoughts in psychology that form the basis for teaching-learning theories that have most influenced the teachinglearning of foreign languages in our culture (Williams & Burden, 1997, Vygotsky, 1934/2007). This study is also based on concepts of language descriptions (Hutchinson & Waters, 1987; Nunan, 1999), principles of Computer-Assisted Language Learning -- CALL (Warschauer & Healey, 1998, 2004, Beatty, 2003; Hardisty & Windeatt, 1989) and the criteria for didactic material evaluation proposed by Ramos (1998). This study, a case study, was conducted with six student-teachers. The data were collected in 2008, by questionnaires. The analysis of the course shows that the language description and the teaching-learning theory that underlie the course Interaction Teachers do not match the paradigms in the Brazilian National Curriculum Parameters to teach foreign languages in elementary and secondary schools (PCN-LE, Brasil, 1998 and PCN-EM, Brazil, 1999) because the course is based on the structural view of language and on the behaviorist teachinglearning theory. As far as the participants perceptions on linguistic and technological aspects exploited in the course are concerned, results show that they have not improved either their linguistic knowledge or technological skills as a consequence of attending the course / Os objetivos deste trabalho são: analisar como as visões de ensinoaprendizagem e linguagem apregoadas pelos PCN-LE (Brasil, 1998) e PCN-EM (Brasil, 1999) se manifestam no curso Interaction Teachers, um curso de inglês mediado por computador, e identificar as percepções de professores-cursistas sobre aspectos linguísticos e tecnológicos desse curso. As bases teóricas desta pesquisa foram: os Parâmetros Curriculares Nacionais Língua Estrangeira (Brasil, 1998) e Parâmetros Curriculares Nacionais Ensino Médio (Brasil, 1999); correntes de pensamento da psicologia que serviram de base para as teorias de ensinoaprendizagem que mais influenciaram o ensino-aprendizagem de língua estrangeira em nossa cultura (Williams & Burden, 1997; Vygotsky, 1934/2007); as teorias de linguagem (Hutchinson & Waters, 1987; Nunan, 1999); o conceito de Computer- Assisted Language Learning CALL (Warschauer & Healey, 1998, 2004; Beatty, 2003; Hardisty & Windeatt, 1989) e os critérios para avaliação de material didático propostos por Ramos (1998). Esta pesquisa, um estudo de caso, foi realizada com seis professores-cursistas. A coleta de dados foi feita no ano de 2008, por meio de questionários. A análise do curso revela que as concepções de linguagem e ensinoaprendizagem que embasam o curso Interaction Teachers contrariam os paradigmas educacionais que subjazem os Parâmetros Curriculares Nacionais Língua Estrangeira (Brasil, 1998) e Parâmetros Curriculares Nacionais Ensino Médio (Brasil, 1999); pois o curso se configura baseado na visão de linguagem estrutural e na visão de ensino-aprendizagem behaviorista. Sobre as percepções a respeito do curso, os professores-cursistas demonstraram que não obtiveram ganhos linguísticos ou tecnológicos como fruto de sua participação no mesmo

published_or_final_version / abstract / toc / Education / Master / Master of Science in Information Technology in Education

This thesis is dedicated to creating a new system with capabilities to create new generic, autonomous strategy computer game controlling elements based on multi-agent systems with social, intelligent decision-making and learning skills. Basic types of strategy games and problems of their playing will be introduced, along with currently used methods of intelligent game AI development. This thesis also presents design and implementation of the new system, working model for a specific game and results obtained while testing it.