In many areas of knowledge, considerable amounts of time have been spent to comprehend and to treat noisy data, one of the most common problems regarding information collection, transmission and storage. These noisy data, when used for training Machine Learning techniques, lead to increased complexity in the induced classification models, higher processing time and reduced predictive power. Treating them in a preprocessing step may improve the data quality and the comprehension of the problem. This Thesis aims to investigate the use of data complexity measures capable to characterize the presence of noise in datasets, to develop new efficient noise ltering techniques in such subsamples of problems of noise identification compared to the state of art and to recommend the most properly suited techniques or ensembles for a specific dataset by meta-learning. Both artificial and real problem datasets were used in the experimental part of this work. They were obtained from public data repositories and a cooperation project. The evaluation was made through the analysis of the effect of artificially generated noise and also by the feedback of a domain expert. The reported experimental results show that the investigated proposals are promising. / Em diversas áreas do conhecimento, um tempo considerável tem sido gasto na compreensão e tratamento de dados ruidosos. Trata-se de uma ocorrência comum quando nos referimos a coleta, a transmissão e ao armazenamento de informações. Esses dados ruidosos, quando utilizados na indução de classificadores por técnicas de Aprendizado de Maquina, aumentam a complexidade da hipótese obtida, bem como o aumento do seu tempo de indução, além de prejudicar sua acurácia preditiva. Trata-los na etapa de pré-processamento pode significar uma melhora da qualidade dos dados e um aumento na compreensão do problema estudado. Esta Tese investiga medidas de complexidade capazes de caracterizar a presença de ruídos em um conjunto de dados, desenvolve novos filtros que sejam mais eficientes em determinados nichos do problema de detecção e remoção de ruídos que as técnicas consideradas estado da arte e recomenda as mais apropriadas técnicas ou comitês de técnicas para um determinado conjunto de dados por meio de meta-aprendizado. As bases de dados utilizadas nos experimentos realizados neste trabalho são tanto artificiais quanto reais, coletadas de repositórios públicos e fornecidas por projetos de cooperação. A avaliação consiste tanto da adição de ruídos artificiais quanto da validação de um especialista. Experimentos realizados mostraram o potencial das propostas investigadas.

In many areas of knowledge, considerable amounts of time have been spent to comprehend and to treat noisy data, one of the most common problems regarding information collection, transmission and storage. These noisy data, when used for training Machine Learning techniques, lead to increased complexity in the induced classification models, higher processing time and reduced predictive power. Treating them in a preprocessing step may improve the data quality and the comprehension of the problem. This Thesis aims to investigate the use of data complexity measures capable to characterize the presence of noise in datasets, to develop new efficient noise ltering techniques in such subsamples of problems of noise identification compared to the state of art and to recommend the most properly suited techniques or ensembles for a specific dataset by meta-learning. Both artificial and real problem datasets were used in the experimental part of this work. They were obtained from public data repositories and a cooperation project. The evaluation was made through the analysis of the effect of artificially generated noise and also by the feedback of a domain expert. The reported experimental results show that the investigated proposals are promising. / Em diversas áreas do conhecimento, um tempo considerável tem sido gasto na compreensão e tratamento de dados ruidosos. Trata-se de uma ocorrência comum quando nos referimos a coleta, a transmissão e ao armazenamento de informações. Esses dados ruidosos, quando utilizados na indução de classificadores por técnicas de Aprendizado de Maquina, aumentam a complexidade da hipótese obtida, bem como o aumento do seu tempo de indução, além de prejudicar sua acurácia preditiva. Trata-los na etapa de pré-processamento pode significar uma melhora da qualidade dos dados e um aumento na compreensão do problema estudado. Esta Tese investiga medidas de complexidade capazes de caracterizar a presença de ruídos em um conjunto de dados, desenvolve novos filtros que sejam mais eficientes em determinados nichos do problema de detecção e remoção de ruídos que as técnicas consideradas estado da arte e recomenda as mais apropriadas técnicas ou comitês de técnicas para um determinado conjunto de dados por meio de meta-aprendizado. As bases de dados utilizadas nos experimentos realizados neste trabalho são tanto artificiais quanto reais, coletadas de repositórios públicos e fornecidas por projetos de cooperação. A avaliação consiste tanto da adição de ruídos artificiais quanto da validação de um especialista. Experimentos realizados mostraram o potencial das propostas investigadas.

abstract: The quality of real-world visual content is typically impaired by many factors including image noise and blur. Detecting and analyzing these impairments are important steps for multiple computer vision tasks. This work focuses on perceptual-based locally adaptive noise and blur detection and their application to image restoration. In the context of noise detection, this work proposes perceptual-based full-reference and no-reference objective image quality metrics by integrating perceptually weighted local noise into a probability summation model. Results are reported on both the LIVE and TID2008 databases. The proposed metrics achieve consistently a good performance across noise types and across databases as compared to many of the best very recent quality metrics. The proposed metrics are able to predict with high accuracy the relative amount of perceived noise in images of different content. In the context of blur detection, existing approaches are either computationally costly or cannot perform reliably when dealing with the spatially-varying nature of the defocus blur. In addition, many existing approaches do not take human perception into account. This work proposes a blur detection algorithm that is capable of detecting and quantifying the level of spatially-varying blur by integrating directional edge spread calculation, probability of blur detection and local probability summation. The proposed method generates a blur map indicating the relative amount of perceived local blurriness. In order to detect the flat/near flat regions that do not contribute to perceivable blur, a perceptual model based on the Just Noticeable Difference (JND) is further integrated in the proposed blur detection algorithm to generate perceptually significant blur maps. We compare our proposed method with six other state-of-the-art blur detection methods. Experimental results show that the proposed method performs the best both visually and quantitatively. This work further investigates the application of the proposed blur detection methods to image deblurring. Two selective perceptual-based image deblurring frameworks are proposed, to improve the image deblurring results and to reduce the restoration artifacts. In addition, an edge-enhanced super resolution algorithm is proposed, and is shown to achieve better reconstructed results for the edge regions. / Dissertation/Thesis / Doctoral Dissertation Electrical Engineering 2016

No description available.

本論文は，5章より構成されている。第1章では，機械学習の現状，応用及び構成を述べた上，本研究で扱った三つの課題を挙げた。第2章では，小サンプルデータの特徴選択方法を提案した。第3章では，クラスの不均衡性と学習データのサイズが分類器精度への影響を検討した。第4章では，ノイズが分類器の学習を妨げる問題点に対して，多要素ベースの学習に基づいた高速クラスノイズの検出方法を提案した。第5章では，分析の主な結果をまとめ，今後の課題と展望を述べた。 / This thesis is organized under five chapters. Chapter 1 gives a brief explanation of what machine learning is and why it matters. Chapter 2 makes a proposal to improve the performance of feature selection methods with low-sample-size data. Chapter 3 studies the effects of class imbalance and training data size on classifier learning empirically. Chapter 4 proposes a fast noise detector referring to the problems of noise detection algorithms, which are over-cleansing, large computational complexity and long response time. Chapter 5 draws a summary and the closing. / 博士(文化情報学) / Doctor of Culture and Information Science / 同志社大学 / Doshisha University

Submitted by Renata Lopes (renatasil82@gmail.com) on 2017-04-20T12:09:28Z No. of bitstreams: 1 thiagorodriguesoliveira.pdf: 2429204 bytes, checksum: b1904f2bde31b546890c5bfa77d58c80 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2017-04-20T12:44:42Z (GMT) No. of bitstreams: 1 thiagorodriguesoliveira.pdf: 2429204 bytes, checksum: b1904f2bde31b546890c5bfa77d58c80 (MD5) / Made available in DSpace on 2017-04-20T12:44:42Z (GMT). No. of bitstreams: 1 thiagorodriguesoliveira.pdf: 2429204 bytes, checksum: b1904f2bde31b546890c5bfa77d58c80 (MD5) Previous issue date: 2010-09-28 / Esta dissertação apresenta, de forma detalhada, um conjunto de metodologias e técnicas destinadas à análise de redes de energia elétrica como meio de transmissão de dados (power line communication - PLC). As características das redes elétricas que influenciam um sistema de comunicação de dados consideradas neste trabalho são as seguintes: a impedância de acesso à rede elétrica, a resposta ao impulso e o ruído. Para tanto, técnicas de processamento de sinais para estimação da resposta em frequência, estimação do comprimento efetivo da resposta ao impulso, detecção e segmentação de ruídos impulsivos e análise espectral de ruídos aditivos são propostas e discutidas na presente contribuição. Os desempenhos objetivos e a apreciação subjetiva das técnicas propostas, a partir de dados sintéticos e medidos, evidenciam a adequação destas técnicas para a análise em questão. Além disso, formulações matemáticas para a resposta ao impulso de canais PLC invariantes, variantes e periodicamente variantes no tempo, derivadas a partir do modelo de multi-propagação para canais PLC, são apresentadas. Tais formulações proporcionam de forma simples e objetiva a emulação dos possíveis comportamentos temporais de canais PLC reais e, portanto, podem se constituir como ferramentas de grande utilidade para o projeto e a avaliação de sistemas de comunicações baseados na tecnologia PLC. / This thesis addresses a set of methodologies and techniques for the analysis of electric grids as a medium for data communications (power line communications - PLC). The main features influencing a communication system that are considered in this work are the input impedance, the channel impulse response, and the noise. In this regards, signal processing-based techniques are investigated, proposed and analyzed for the estimations of the channel frequency response and the effective length of the channel impulse response; the detection and segmentation of impulsive noise; and the power spectral analysis of the additive noise at the channel output. The numerical performance and subjective analysis regarding the use of the proposed techniques in synthetic and measured data indicate that those techniques fit well in the thesis purposes. In addition, mathematical formulation for invariant, time-varying, and periodically time-varying PLC channel models, which are based on multi-path channel model approach, are presented. These formulations are simple and elegant ones for the emulation of possible temporal behavior of existing PLC channels and, as a result, can constitute a useful tool for the design and analysis of PLC systems.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2017-04-24T15:28:35Z No. of bitstreams: 1 paolajohanasaboyalopez.pdf: 1042873 bytes, checksum: a46dd95de00e062cba39ef4b9b642462 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2017-04-24T17:09:24Z (GMT) No. of bitstreams: 1 paolajohanasaboyalopez.pdf: 1042873 bytes, checksum: a46dd95de00e062cba39ef4b9b642462 (MD5) / Made available in DSpace on 2017-04-24T17:09:24Z (GMT). No. of bitstreams: 1 paolajohanasaboyalopez.pdf: 1042873 bytes, checksum: a46dd95de00e062cba39ef4b9b642462 (MD5) Previous issue date: 2013-06-03 / A presente dissertação tem por objetivo propor e avaliar cinco técnicas de detecção de ruídos impulsivos para a melhoria da transmissão digital de dados via redes de energia elétrica (do inglês, Power Line Communications) (PLC). As técnicas propostas contemplam a detecção de ruídos impulsivos no domínio do tempo discreto, no domínio da transformada wavelet discreta (do inglês, Discrete Wavelet Transform) (DWT) e no domínio da transformada discreta de Fourier (do inglês, Discrete Fourier Transform) (DFT). Tais técnicas fazem uso de métodos de extração e seleção de características, assim como métodos de detecção de sinais baseados na teoria de Bayes e redes neurais. Análises comparativas explicitam as vantagens e desvantagens de cada uma das técnicas propostas para o problema em questão, e ainda indicam que estas são bastante adequadas para a solução do mesmo. / This dissertation aims to propose and evaluate five techniques for impulsive noise detection in order to improve digital communications through power line channels. The imput signals for the proposed detection techniques are impulsive noise signals on discrete-time domain, on the Discrete Wavelet Transform domain and on the Discrete Fourier Transform domain and it makes use of feature extraction and selection techniques, as well as detection techniques supported on Bayes Theory and Multi-layer Perceptron Neural Networks. Comparative analysis show some advantages and disadvantages of each proposed technique and the relevance of them to solve the impulsive noise detection problem.