Representació no-lineal de les imatges per a codificació perceptiva

JPEG2000 és un estàndard de compressió d’imatges que utilitza la transformada wavelet i, posteriorment, una quantificació uniforme dels coeficients amb dead-zone. Els coeficients wavelet presenten certes dependències tant estadístiques com visuals. Les dependències estadístiques es tenen en compte a l'esquema JPEG2000, no obstant, no passa el mateix amb les dependències visuals. En aquest treball, es pretén trobar una representació més adaptada al sistema visual que la que proporciona JPEG2000 directament. Per trobar-la utilitzarem la normalització divisiva dels coeficients, tècnica que ja ha demostrat resultats tant en decorrelació estadística de coeficients com perceptiva. Idealment, el que es voldria fer és reconvertir els coeficients a un espai de valors en els quals un valor més elevat dels coeficients impliqui un valor més elevat d'aportació visual, i utilitzar aquest espai de valors per a codificar. A la pràctica, però, volem que el nostre sistema de codificació estigui integrat a un estàndard. És per això que utilitzarem JPEG2000, estàndard de la ITU que permet una elecció de les distorsions en la codificació, i utilitzarem la distorsió en el domini de coeficients normalitzats com a mesura de distorsió per a escollir quines dades s'envien abans.

trEST, trGEN and Hits: access to databases of predicted protein sequences.

High throughput genome (HTG) and expressed sequence tag (EST) sequences are currently the most abundant nucleotide sequence classes in the public database. The large volume, high degree of fragmentation and lack of gene structure annotations prevent efficient and effective searches of HTG and EST data for protein sequence homologies by standard search methods. Here, we briefly describe three newly developed resources that should make discovery of interesting genes in these sequence classes easier in the future, especially to biologists not having access to a powerful local bioinformatics environment. trEST and trGEN are regularly regenerated databases of hypothetical protein sequences predicted from EST and HTG sequences, respectively. Hits is a web-based data retrieval and analysis system providing access to precomputed matches between protein sequences (including sequences from trEST and trGEN) and patterns and profiles from Prosite and Pfam. The three resources can be accessed via the Hits home page (http://hits. isb-sib.ch).

Processament de música basat en el seu contingut

Projecte de recerca elaborat a partir d’una estada a la National University of Singapore Singapur, entre juliol i octubre del 2007. Donada l'explosió de la música a l'internet i la ràpida expansió de les col•leccions de música digital, un repte clau en l'àrea de la informació musical és el desenvolupament de sistemes de processament musical eficients i confiables. L'objectiu de la investigació proposada ha estat treballar en diferents aspectes de l'extracció, modelatge i processat del contingut musical. En particular, s’ha treballat en l'extracció, l'anàlisi i la manipulació de descriptors d'àudio de baix nivell, el modelatge de processos musicals, l'estudi i desenvolupament de tècniques d'aprenentatge automàtic per a processar àudio, i la identificació i extracció d'atributs musicals d'alt nivell. S’han revisat i millorat alguns components d'anàlisis d'àudio i revisat components per a l'extracció de descriptors inter-nota i intra-nota en enregistraments monofónics d'àudio. S’ha aplicat treball previ en Tempo a la formalització de diferents tasques musicals. Finalment, s’ha investigat el processat d'alt nivell de música basandonos en el seu contingut. Com exemple d'això, s’ha investigat com músics professionals expressen i comuniquen la seva interpretació del contingut musical i emocional de peces musicals, i hem usat aquesta informació per a identificar automàticament intèrprets. S’han estudiat les desviacions en paràmetres com to, temps, amplitud i timbre a nivell inter-nota i intra-nota.

Usefulness of microsatellite typing in population genetic studies of Trypanosoma cruzi

Through microsatellite analysis of 53 monoclonal populations of Trypanosoma cruzi, we found a remarkable degree of genetic polymorphism with no single multilocus genotype being observed more than once. The microsatellite profile proved to be stable during 70 generations of the CL Brener clone in culture. The microsatellite profiling presented also high diagnostic sensitivity since DNA amplifications could be achieved with less than 100 fg DNA, corresponding to half parasite total DNA content. Based on these technical attributes the microsatellite assay turns out to be an important tool for direct typing T. cruzi in biological samples. By using this approach we were able to type T. cruzi in feces of artificially infected bugs and in single cells sorted by FACS. The microsatellites have shown to be excellent markers for T. cruzi phylogenetic reconstruction. We used maximum parsimony based on the minimum number of mutational steps to build an unrooted Wagner network, which confirms previous conclusions based on the analysis of the D7 domain of the LSU rDNA gene that T. cruzi is composed by two major groups. We also obtained evidence that strains belonging to rRNA group 2 are subdivided into two genetically distant clusters, and that one of these clusters is more related to rRNA group 1/2. These results suggest different origins for these strains.

Scene Classification Using a Hybrid Generative/Discriminative Approach

We investigate whether dimensionality reduction using a latent generative model is beneficial for the task of weakly supervised scene classification. In detail, we are given a set of labeled images of scenes (for example, coast, forest, city, river, etc.), and our objective is to classify a new image into one of these categories. Our approach consists of first discovering latent ";topics"; using probabilistic Latent Semantic Analysis (pLSA), a generative model from the statistical text literature here applied to a bag of visual words representation for each image, and subsequently, training a multiway classifier on the topic distribution vector for each image. We compare this approach to that of representing each image by a bag of visual words vector directly and training a multiway classifier on these vectors. To this end, we introduce a novel vocabulary using dense color SIFT descriptors and then investigate the classification performance under changes in the size of the visual vocabulary, the number of latent topics learned, and the type of discriminative classifier used (k-nearest neighbor or SVM). We achieve superior classification performance to recent publications that have used a bag of visual word representation, in all cases, using the authors' own data sets and testing protocols. We also investigate the gain in adding spatial information. We show applications to image retrieval with relevance feedback and to scene classification in videos

Reporting intellectual capital in Spain

This study reports on the analysis of annual reports from 14- listed companies in Spainover a five-year period, from 1998 to 2002. Companies in the sample are selected on thebasis of their knowledge-based assets and incentives to report on Intellectual Capital.The empirical analysis is twofold:1) Firstly, we analyse the value of intellectual capital using a value-based approach,through the difference between market and book value over the period considered. Results show that there is a general decrease in the 'hidden value' of these companies, probably due to the general trend in stock markets.2) Secondly, we carry out a content-based analysis of the complete annual reports of the companies over the five year period. Preliminary findings seem to suggest that although the level of disclosure has increased over time, this is mainly in the form of narrative. Overall, the level of disclosure of intellectual capital remains low.

CLIL Materials in Chemistry and English: Iodometric Titrations

Content and Language Integrated Learning (CLIL)Materials in Chemistry and English following the principles of CLIL / Content-based Instruction and Task-based Learning

CLIL Materials in Chemistry and English: Iodometric Titrations

Content and Language Integrated Learning (CLIL)Materials in Chemistry and English following the principles of CLIL / Content-based Instruction and Task-based Learning

Hand detection in cluttered scene images using Fourier-Mellin invariant features

This paper proposes an automatic hand detection system that combines the Fourier-Mellin Transform along with other computer vision techniques to achieve hand detection in cluttered scene color images. The proposed system uses the Fourier-Mellin Transform as an invariant feature extractor to perform RST invariant hand detection. In a first stage of the system a simple non-adaptive skin color-based image segmentation and an interest point detector based on corners are used in order to identify regions of interest that contains possible matches. A sliding window algorithm is then used to scan the image at different scales performing the FMT calculations only in the previously detected regions of interest and comparing the extracted FM descriptor of the windows with a hand descriptors database obtained from a train image set. The results of the performed experiments suggest the use of Fourier-Mellin invariant features as a promising approach for automatic hand detection.

Hand detection in cluttered scene images using Fourier-Mellin invariant features

This paper proposes an automatic hand detection system that combines the Fourier-Mellin Transform along with other computer vision techniques to achieve hand detection in cluttered scene color images. The proposed system uses the Fourier-Mellin Transform as an invariant feature extractor to perform RST invariant hand detection. In a first stage of the system a simple non-adaptive skin color-based image segmentation and an interest point detector based on corners are used in order to identify regions of interest that contains possible matches. A sliding window algorithm is then used to scan the image at different scales performing the FMT calculations only in the previously detected regions of interest and comparing the extracted FM descriptor of the windows with a hand descriptors database obtained from a train image set. The results of the performed experiments suggest the use of Fourier-Mellin invariant features as a promising approach for automatic hand detection.

Learning journalistic opinion genres using CLIL methodology

The Faculty of Business and Communication recently started an internationalization process that, in two year’s time, will allow all undergraduate students (studying Journalism, Audiovisual Communication, Advertising and Public Relations, Business and Marketing) to take 25% of their subjects in English using CLIL methodology. Currently, Journalism is the degree course with the greatest percentage of CLIL subjects, for example Current Affairs Workshop, a subject dedicated to analyzing current news using opinion genres. Moreover, because of the lack of other subjects offered in English, ERASMUS students have to take some journalism subjects in order to complete their international passport, and one of the classes they choose is the Current Affairs Workshop. The aim of this paper is to explore how CLIL methodology can be useful for learning journalistic opinion genres (chat-shows, discussions and debates) in a subject where Catalan Communication students –with different levels of English- share their knowledge with European students of other social disciplines. Students work in multidisciplinary groups in which they develop real radio and TV programs, adopting all the roles (moderator, technician, producer and participants), analyzing daily newspapers and other sources to create content, based on current affairs. This paper is based on the participant observation of the lecturers of the subject, who have designed different activities related to journalistic genres, where students can develop their skills according to the role they play in every assignment. Examples of successful lessons will be given, in addition to the results of the course: both positive and negative. Although the objective of the course is to examine professional routines related to opinion genres, and students are not directly graded on their level of English, the Catalan students come to appreciate how they finally overcome their fear of working in a foreign language. This is a basic result of their experience.

Creating and organizing an "English class" in primary education: The role of teacher talk in the EFL amd CLIL Classes

Catalonia is a bilingual country where the presence of English in the social context is small; the amount of input received by the primary education pupils is very little and this input mainly comes from the English lessons at school. Consequently, this situation combined with the increasing demand for English and the fact that the new generations want to become communicatively competent in English place the role of English teachers in a relevant position. This research project analyses the role of the English teacher talk; in particular, the study focuses on the teacher’s oral productions in foreign language lessons (EFL) and in content-based lessons (CLIL).

Metadata for describing learning scenarios under European Higher Education Area paradigm

In this paper we identify the requirements for creating formal descriptions of learning scenarios designed under the European HigherEducation Area paradigm, using competences and learning activities as the basic pieces of the learning process, instead of contents and learning resources, pursuing personalization. Classical arrangements of content based courses are no longer enough to describe all the richness of this new learning process, where user profiles, competences and complex hierarchical itineraries need to be properly combined. We study the intersection with the current IMS Learning Design specification and theadditional metadata required for describing such learning scenarios. This new approach involves the use of case based learning and collaborativelearning in order to acquire and develop competences, following adaptive learning paths in two structured levels.

Informed recommender: basing recommendations on consumer product reviews

Recommender systems attempt to predict items in which a user might be interested, given some information about the user's and items' profiles. Most existing recommender systems use content-based or collaborative filtering methods or hybrid methods that combine both techniques (see the sidebar for more details). We created Informed Recommender to address the problem of using consumer opinion about products, expressed online in free-form text, to generate product recommendations. Informed recommender uses prioritized consumer product reviews to make recommendations. Using text-mining techniques, it maps each piece of each review comment automatically into an ontology

Analysis of Satellite Data for Forest Segmentation

In this work we study the classification of forest types using mathematics based image analysis on satellite data. We are interested in improving classification of forest segments when a combination of information from two or more different satellites is used. The experimental part is based on real satellite data originating from Canada. This thesis gives summary of the mathematics basics of the image analysis and supervised learning , methods that are used in the classification algorithm. Three data sets and four feature sets were investigated in this thesis. The considered feature sets were 1) histograms (quantiles) 2) variance 3) skewness and 4) kurtosis. Good overall performances were achieved when a combination of ASTERBAND and RADARSAT2 data sets was used.