Aplicació de visió per dispositius mòbils

En aquest projecte es presenta l’aplicació per a dispositius mòbils Doppelganger. La seva funció és, a partir d’una fotografia, detectar la cara i mostrar la persona famosa de la nostra base de dades que més s’assembla a la persona en la fotografia. Per la implementació s’han utilitzat algoritmes de visió per computador i d’aprenentatge automàtic com per exemple el PCA i el K-Nearest Neighbor, tot utilitzant llibreries gratuïtes com són les OpenCV.

Predicting clinical scores from magnetic resonance scans in Alzheimer's disease.

Machine learning and pattern recognition methods have been used to diagnose Alzheimer's disease (AD) and mild cognitive impairment (MCI) from individual MRI scans. Another application of such methods is to predict clinical scores from individual scans. Using relevance vector regression (RVR), we predicted individuals' performances on established tests from their MRI T1 weighted image in two independent data sets. From Mayo Clinic, 73 probable AD patients and 91 cognitively normal (CN) controls completed the Mini-Mental State Examination (MMSE), Dementia Rating Scale (DRS), and Auditory Verbal Learning Test (AVLT) within 3months of their scan. Baseline MRI's from the Alzheimer's disease Neuroimaging Initiative (ADNI) comprised the other data set; 113 AD, 351 MCI, and 122 CN subjects completed the MMSE and Alzheimer's Disease Assessment Scale-Cognitive subtest (ADAS-cog) and 39 AD, 92 MCI, and 32 CN ADNI subjects completed MMSE, ADAS-cog, and AVLT. Predicted and actual clinical scores were highly correlated for the MMSE, DRS, and ADAS-cog tests (P<0.0001). Training with one data set and testing with another demonstrated stability between data sets. DRS, MMSE, and ADAS-Cog correlated better than AVLT with whole brain grey matter changes associated with AD. This result underscores their utility for screening and tracking disease. RVR offers a novel way to measure interactions between structural changes and neuropsychological tests beyond that of univariate methods. In clinical practice, we envision using RVR to aid in diagnosis and predict clinical outcome.

Multispectral image compression using the wavelet transform

Multispectral images contain information from several spectral wavelengths and currently multispectral images are widely used in remote sensing and they are becoming more common in the field of computer vision and in industrial applications. Typically, one multispectral image in remote sensing may occupy hundreds of megabytes of disk space and several this kind of images may be received from a single measurement. This study considers the compression of multispectral images. The lossy compression is based on the wavelet transform and we compare the suitability of different waveletfilters for the compression. A method for selecting a wavelet filter for the compression and reconstruction of multispectral images is developed. The performance of the multidimensional wavelet transform based compression is compared to other compression methods like PCA, ICA, SPIHT, and DCT/JPEG. The quality of the compression and reconstruction is measured by quantitative measures like signal-to-noise ratio. In addition, we have developed a qualitative measure, which combines the information from the spatial and spectral dimensions of a multispectral image and which also accounts for the visual quality of the bands from the multispectral images.

Similarity measure based classification

Luokittelujärjestelmää suunniteltaessa tarkoituksena on rakentaa systeemi, joka pystyy ratkaisemaan mahdollisimman tarkasti tutkittavan ongelma-alueen. Hahmontunnistuksessa tunnistusjärjestelmän ydin on luokitin. Luokittelun sovellusaluekenttä on varsin laaja. Luokitinta tarvitaan mm. hahmontunnistusjärjestelmissä, joista kuvankäsittely toimii hyvänä esimerkkinä. Myös lääketieteen parissa tarkkaa luokittelua tarvitaan paljon. Esimerkiksi potilaan oireiden diagnosointiin tarvitaan luokitin, joka pystyy mittaustuloksista päättelemään mahdollisimman tarkasti, onko potilaalla kyseinen oire vai ei. Väitöskirjassa on tehty similaarisuusmittoihin perustuva luokitin ja sen toimintaa on tarkasteltu mm. lääketieteen paristatulevilla data-aineistoilla, joissa luokittelutehtävänä on tunnistaa potilaan oireen laatu. Väitöskirjassa esitetyn luokittimen etuna on sen yksinkertainen rakenne, josta johtuen se on helppo tehdä sekä ymmärtää. Toinen etu on luokittimentarkkuus. Luokitin saadaan luokittelemaan useita eri ongelmia hyvin tarkasti. Tämä on tärkeää varsinkin lääketieteen parissa, missä jo pieni tarkkuuden parannus luokittelutuloksessa on erittäin tärkeää. Väitöskirjassa ontutkittu useita eri mittoja, joilla voidaan mitata samankaltaisuutta. Mitoille löytyy myös useita parametreja, joille voidaan etsiä juuri kyseiseen luokitteluongelmaan sopivat arvot. Tämä parametrien optimointi ongelma-alueeseen sopivaksi voidaan suorittaa mm. evoluutionääri- algoritmeja käyttäen. Kyseisessä työssä tähän on käytetty geneettistä algoritmia ja differentiaali-evoluutioalgoritmia. Luokittimen etuna on sen joustavuus. Ongelma-alueelle on helppo vaihtaa similaarisuusmitta, jos kyseinen mitta ei ole sopiva tutkittavaan ongelma-alueeseen. Myös eri mittojen parametrien optimointi voi parantaa tuloksia huomattavasti. Kun käytetään eri esikäsittelymenetelmiä ennen luokittelua, tuloksia pystytään parantamaan.

Level Set Methods in Image Processing

This thesis gives an overview of the use of the level set methods in the field of image science. The similar fast marching method is discussed for comparison, also the narrow band and the particle level set methods are introduced. The level set method is a numerical scheme for representing, deforming and recovering structures in an arbitrary dimensions. It approximates and tracks the moving interfaces, dynamic curves and surfaces. The level set method does not define how and why some boundary is advancing the way it is but simply represents and tracks the boundary. The principal idea of the level set method is to represent the N dimensional boundary in the N+l dimensions. This gives the generality to represent even the complex boundaries. The level set methods can be powerful tools to represent dynamic boundaries, but they can require lot of computing power. Specially the basic level set method have considerable computational burden. This burden can be alleviated with more sophisticated versions of the level set algorithm like the narrow band level set method or with the programmable hardware implementation. Also the parallel approach can be used in suitable applications. It is concluded that these methods can be used in a quite broad range of image applications, like computer vision and graphics, scientific visualization and also to solve problems in computational physics. Level set methods and methods derived and inspired by it will be in the front line of image processing also in the future.

Apprentissage des réseaux de neurones profonds et applications en traitement automatique de la langue naturelle

En apprentissage automatique, domaine qui consiste à utiliser des données pour apprendre une solution aux problèmes que nous voulons confier à la machine, le modèle des Réseaux de Neurones Artificiels (ANN) est un outil précieux. Il a été inventé voilà maintenant près de soixante ans, et pourtant, il est encore de nos jours le sujet d'une recherche active. Récemment, avec l'apprentissage profond, il a en effet permis d'améliorer l'état de l'art dans de nombreux champs d'applications comme la vision par ordinateur, le traitement de la parole et le traitement des langues naturelles. La quantité toujours grandissante de données disponibles et les améliorations du matériel informatique ont permis de faciliter l'apprentissage de modèles à haute capacité comme les ANNs profonds. Cependant, des difficultés inhérentes à l'entraînement de tels modèles, comme les minima locaux, ont encore un impact important. L'apprentissage profond vise donc à trouver des solutions, en régularisant ou en facilitant l'optimisation. Le pré-entraînnement non-supervisé, ou la technique du ``Dropout'', en sont des exemples. Les deux premiers travaux présentés dans cette thèse suivent cette ligne de recherche. Le premier étudie les problèmes de gradients diminuants/explosants dans les architectures profondes. Il montre que des choix simples, comme la fonction d'activation ou l'initialisation des poids du réseaux, ont une grande influence. Nous proposons l'initialisation normalisée pour faciliter l'apprentissage. Le second se focalise sur le choix de la fonction d'activation et présente le rectifieur, ou unité rectificatrice linéaire. Cette étude a été la première à mettre l'accent sur les fonctions d'activations linéaires par morceaux pour les réseaux de neurones profonds en apprentissage supervisé. Aujourd'hui, ce type de fonction d'activation est une composante essentielle des réseaux de neurones profonds. Les deux derniers travaux présentés se concentrent sur les applications des ANNs en traitement des langues naturelles. Le premier aborde le sujet de l'adaptation de domaine pour l'analyse de sentiment, en utilisant des Auto-Encodeurs Débruitants. Celui-ci est encore l'état de l'art de nos jours. Le second traite de l'apprentissage de données multi-relationnelles avec un modèle à base d'énergie, pouvant être utilisé pour la tâche de désambiguation de sens.

A Comparative Study of Wavelet Based Feature Extraction Techniques in Recognizing Isolated Spoken Words

Speech is a natural mode of communication for people and speech recognition is an intensive area of research due to its versatile applications. This paper presents a comparative study of various feature extraction methods based on wavelets for recognizing isolated spoken words. Isolated words from Malayalam, one of the four major Dravidian languages of southern India are chosen for recognition. This work includes two speech recognition methods. First one is a hybrid approach with Discrete Wavelet Transforms and Artificial Neural Networks and the second method uses a combination of Wavelet Packet Decomposition and Artificial Neural Networks. Features are extracted by using Discrete Wavelet Transforms (DWT) and Wavelet Packet Decomposition (WPD). Training, testing and pattern recognition are performed using Artificial Neural Networks (ANN). The proposed method is implemented for 50 speakers uttering 20 isolated words each. The experimental results obtained show the efficiency of these techniques in recognizing speech

Molecular genetic profiling of functional genes of pearl oyster, Pinctada fucata

There are a number of genes involved in the regulation of functional process in marine bivalves. In the case of pearl oyster, some of these genes have major role in the immune/defence function and biomineralization process involved in the pearl formation in them. As secondary filter feeders, pearl oysters are exposed to various kinds of stressors like bacteria, viruses, pesticides, industrial wastes, toxic metals and petroleum derivatives, making susceptible to diseases. Environmental changes and ambient stress also affect non-specific immunity, making the organisms vulnerable to infections. These stressors can trigger various cellular responses in the animals in their efforts to counteract the ill effects of the stress on them. These include the expression of defence related genes which encode factors such as antioxidant genes, pattern recognition receptor proteins etc. One of the strategies to combat these problems is to get insight into the disease resistance genes, and use them for disease control and health management. Similarly, although it is known that formation of pearl in molluscs is mediated by specialized proteins which are in turn regulated by specific genes encoding them, there is a paucity of sufficient information on these genes.In view of the above facts, studies on the defence related and pearl forming genes of the pearl oyster assumes importance from the point of view of both sustainable fishery management and aquaculture. At present, there is total lack of sufficient knowledge on the functional genes and their expressions in the Indian pearl oyster Pinctada fucata. Hence this work was taken up to identify and characterize the defence related and pearl forming genes, and study their expression through molecular means, in the Indian pearl oyster Pinctada fucata which are economically important for aquaculture at the southeast coast of India. The present study has successfully carried out the molecular identification, characterization and expression analysis of defence related antioxidant enzyme genes and pattern recognition proteins genes which play vital role in the defence against biotic and abiotic stressors. Antioxidant enzyme genes viz., Cu/Zn superoxide dismutase (Cu/Zn SOD), glutathione peroxidise (GPX) and glutathione-S-transferase (GST) were studied. Concerted approaches using the various molecular tools like polymerase chain reaction (PCR), random amplification of cDNA ends (RACE), molecular cloning and sequencing have resulted in the identification and characterization of full length sequences (924 bp) of the Cu/Zn SOD, most important antioxidant enzyme gene. BLAST search in NCBI confirmed the identity of the gene as Cu/Zn SOD. The presence of the characteristic amino acid sequences such as copper/zinc binding residues, family signature sequences and signal peptides were found out. Multiple sequence alignment comparison and phylogenetic analysis of the nucleotide and amino acid sequences using bioinformatics tools like BioEdit,MEGA etc revealed that the sequences were found to contain regions of diversity as well as homogeneity. Close evolutionary relationship between P. fucata and other aquatic invertebrates was revealed from the phylogenetic tree constructed using SOD amino acid sequence of P. fucata and other invertebrates as well as vertebrates

Eigen Faces: Detecció, Classificació/Reconeixement de cares

En aquest projecte es pretén utilitzar mètodes coneguts com ara Viola&Jones (detecció) i EigenFaces (reconeixement) per a detectar i reconèixer cares dintre d’imatges de vídeo. Per a aconseguir aquesta tasca cal partir d’un conjunt de dades d’entrenament per a cada un dels mètodes (base de dades formada per imatges i anotacions manuals). A partir d’aquí, l’aplicació, ha de ser capaç de detectar cares en noves imatges i reconèixer-les (identificar de quina cara es tracta)

A survey addressing the fundamental matrix estimation problem

Epipolar geometry is a key point in computer vision and the fundamental matrix estimation is the only way to compute it. This article surveys several methods of fundamental matrix estimation which have been classified into linear methods, iterative methods and robust methods. All of these methods have been programmed and their accuracy analysed using real images. A summary, accompanied with experimental results, is given

VLSI architecture for motion estimation in underwater imaging

El treball desenvolupat en aquesta tesi aprofundeix i aporta solucions innovadores en el camp orientat a tractar el problema de la correspondència en imatges subaquàtiques. En aquests entorns, el que realment complica les tasques de processat és la falta de contorns ben definits per culpa d'imatges esborronades; un fet aquest que es deu fonamentalment a il·luminació deficient o a la manca d'uniformitat dels sistemes d'il·luminació artificials. Els objectius aconseguits en aquesta tesi es poden remarcar en dues grans direccions. Per millorar l'algorisme d'estimació de moviment es va proposar un nou mètode que introdueix paràmetres de textura per rebutjar falses correspondències entre parells d'imatges. Un seguit d'assaigs efectuats en imatges submarines reals han estat portats a terme per seleccionar les estratègies més adients. Amb la finalitat d'aconseguir resultats en temps real, es proposa una innovadora arquitectura VLSI per la implementació d'algunes parts de l'algorisme d'estimació de moviment amb alt cost computacional.

An overview of the use of neural networks for data mining tasks

In the recent years, the area of data mining has been experiencing considerable demand for technologies that extract knowledge from large and complex data sources. There has been substantial commercial interest as well as active research in the area that aim to develop new and improved approaches for extracting information, relationships, and patterns from large datasets. Artificial neural networks (NNs) are popular biologically-inspired intelligent methodologies, whose classification, prediction, and pattern recognition capabilities have been utilized successfully in many areas, including science, engineering, medicine, business, banking, telecommunication, and many other fields. This paper highlights from a data mining perspective the implementation of NN, using supervised and unsupervised learning, for pattern recognition, classification, prediction, and cluster analysis, and focuses the discussion on their usage in bioinformatics and financial data analysis tasks. © 2012 Wiley Periodicals, Inc.

Biologically-inspired robust motion segmentation using mutual information

This paper presents a neuroscience inspired information theoretic approach to motion segmentation. Robust motion segmentation represents a fundamental first stage in many surveillance tasks. As an alternative to widely adopted individual segmentation approaches, which are challenged in different ways by imagery exhibiting a wide range of environmental variation and irrelevant motion, this paper presents a new biologically-inspired approach which computes the multivariate mutual information between multiple complementary motion segmentation outputs. Performance evaluation across a range of datasets and against competing segmentation methods demonstrates robust performance.

Monitoring vegetation on railway embankments : supporting maintenance decisions

The national railway administrations in Scandinavia, Germany, and Austria mainly resort to manual inspections to control vegetation growth along railway embankments. Manually inspecting railways is slow and time consuming. A more worrying aspect concerns the fact that human observers are often unable to estimate the true cover of vegetation on railway embankments. Further human observers often tend to disagree with each other when more than one observer is engaged for inspection. Lack of proper techniques to identify the true cover of vegetation even result in the excess usage of herbicides; seriously harming the environment and threating the ecology. Hence work in this study has investigated aspects relevant to human variationand agreement to be able to report better inspection routines. This was studied by mainly carrying out two separate yet relevant investigations.First, thirteen observers were separately asked to estimate the vegetation cover in nine imagesacquired (in nadir view) over the railway tracks. All such estimates were compared relatively and an analysis of variance resulted in a significant difference on the observers’ cover estimates (p<0.05). Bearing in difference between the observers, a second follow-up field-study on the railway tracks was initiated and properly investigated. Two railway segments (strata) representingdifferent levels of vegetationwere carefully selected. Five sample plots (each covering an area of one-by-one meter) were randomizedfrom each stratumalong the rails from the aforementioned segments and ten images were acquired in nadir view. Further three observers (with knowledge in the railway maintenance domain) were separately asked to estimate the plant cover by visually examining theplots. Again an analysis of variance resulted in a significant difference on the observers’ cover estimates (p<0.05) confirming the result from the first investigation.The differences in observations are compared against a computer vision algorithm which detects the "true" cover of vegetation in a given image. The true cover is defined as the amount of greenish pixels in each image as detected by the computer vision algorithm. Results achieved through comparison strongly indicate that inconsistency is prevalent among the estimates reported by the observers. Hence, an automated approach reporting the use of computer vision is suggested, thus transferring the manual inspections into objective monitored inspections

Técnicas de computação natural para segmentação de imagens médicas

Image segmentation is one of the image processing problems that deserves special attention from the scientific community. This work studies unsupervised methods to clustering and pattern recognition applicable to medical image segmentation. Natural Computing based methods have shown very attractive in such tasks and are studied here as a way to verify it's applicability in medical image segmentation. This work treats to implement the following methods: GKA (Genetic K-means Algorithm), GFCMA (Genetic FCM Algorithm), PSOKA (PSO and K-means based Clustering Algorithm) and PSOFCM (PSO and FCM based Clustering Algorithm). Besides, as a way to evaluate the results given by the algorithms, clustering validity indexes are used as quantitative measure. Visual and qualitative evaluations are realized also, mainly using data given by the BrainWeb brain simulator as ground truth