Maximum likelihood Linear Programming Data Fusion for Speaker Recognition

Biometric system performance can be improved by means of data fusion. Several kinds of information can be fused in order to obtain a more accurate classification (identification or verification) of an input sample. In this paper we present a method for computing the weights in a weighted sum fusion for score combinations, by means of a likelihood model. The maximum likelihood estimation is set as a linear programming problem. The scores are derived from a GMM classifier working on a different feature extractor. Our experimental results assesed the robustness of the system in front a changes on time (different sessions) and robustness in front a change of microphone. The improvements obtained were significantly better (error bars of two standard deviations) than a uniform weighted sum or a uniform weighted product or the best single classifier. The proposed method scales computationaly with the number of scores to be fussioned as the simplex method for linear programming.

Non-Linear and Non-Conventional Speech Processing: Alternative Techniques

This special issue aims to cover some problems related to non-linear and nonconventional speech processing. The origin of this volume is in the ISCA Tutorial and Research Workshop on Non-Linear Speech Processing, NOLISP’09, held at the Universitat de Vic (Catalonia, Spain) on June 25–27, 2009. The series of NOLISP workshops started in 2003 has become a biannual event whose aim is to discuss alternative techniques for speech processing that, in a sense, do not fit into mainstream approaches. A selected choice of papers based on the presentations delivered at NOLISP’09 has given rise to this issue of Cognitive Computation.

Empirical mode decomposition-based face recognition system

In this work we explore the multivariate empirical mode decomposition combined with a Neural Network classifier as technique for face recognition tasks. Images are simultaneously decomposed by means of EMD and then the distance between the modes of the image and the modes of the representative image of each class is calculated using three different distance measures. Then, a neural network is trained using 10- fold cross validation in order to derive a classifier. Preliminary results (over 98 % of classification rate) are satisfactory and will justify a deep investigation on how to apply mEMD for face recognition.

Plagiarism meets paraphrasing: insights for the new generation in automatic plagiarism detection

Although paraphrasing is the linguistic mechanism underlying many plagiarism cases, little attention has been paid to its analysis in the framework of automatic plagiarism detection. Therefore, state-of-the-art plagiarism detectors find it difficult to detect cases of paraphrase plagiarism. In this article, we analyse the relationship between paraphrasing and plagiarism, paying special attention to which paraphrase phenomena underlie acts of plagiarism and which of them are detected by plagiarism detection systems. With this aim in mind, we created the P4P corpus, a new resource which uses a paraphrase typology to annotate a subset of the PAN-PC-10 corpus for automatic plagiarism detection. The results of the Second International Competition on Plagiarism Detection were analysed in the light of this annotation. The presented experiments show that (i) more complex paraphrase phenomena and a high density of paraphrase mechanisms make plagiarism detection more difficult, (ii) lexical substitutions are the paraphrase mechanisms used the most when plagiarising, and (iii) paraphrase mechanisms tend to shorten the plagiarized text. For the first time, the paraphrase mechanisms behind plagiarism have been analysed, providing critical insights for the improvement of automatic plagiarism detection systems.

Auditory localisation and recognition after left inferior collicular lesion: effects of work load on cortical activation patterns (fMRI study)

Evidence from neuropsychological and activation studies (Clarke et al., 2oo0, Maeder et al., 2000) suggests that sound recognitionand localisation are processed by two anatomically and functionally distinct cortical networks. We report here on a case of a patientthat had an interruption of auditory information and we show: i) the effects of this interruption on cortical auditory processing; ii)the effect of the workload on activation pattern.A 36 year old man suffered from a small left mesencephalic haemotrhage, due to cavernous angioma; the let% inferior colliculuswas resected in the surgical approach of the vascular malformation. In the acute stage, the patient complained of auditoryhallucinations and of auditory loss in right ear, while tonal audiometry was normal. At 12 months, auditory recognition, auditorylocalisation (assessed by lTD and IID cues) and auditory motion perception were normal (Clarke et al., 2000), while verbal dichoticlistening was deficient on the right side.Sound recognition and sound localisation activation patterns were investigated with fMRI, using a passive and an activeparadigm. In normal subjects, distinct cortical networks were involved in sound recognition and localisation, both in passive andactive paradigm (Maeder et al., 2OOOa, 2000b).Passive listening of environmental and spatial stimuli as compared to rest strongly activated right auditory cortex, but failed toactivate left primary auditory cortex. The specialised networks for sound recognition and localisation could not be visual&d onthe right and only minimally on the left convexity. A very different activation pattern was obtained in the active condition wherea motor response was required. Workload not only increased the activation of the right auditory cortex, but also allowed theactivation of the left primary auditory cortex. The specialised networks for sound recognition and localisation were almostcompletely present in both hemispheres.These results show that increasing the workload can i) help to recruit cortical region in the auditory deafferented hemisphere;and ii) lead to processing auditory information within specific cortical networks.References:Clarke et al. (2000). Neuropsychologia 38: 797-807.Mae.der et al. (2OOOa), Neuroimage 11: S52.Maeder et al. (2OOOb), Neuroimage 11: S33

Computed tomography commissioning programmes: how to obtain a reliable MTF with an automatic approach?

The purpose of this study was to assess the spatial resolution of a computed tomography (CT) scanner with an automatic approach developed for routine quality controls when varying CT parameters. The methods available to assess the modulation transfer functions (MTF) with the automatic approach were Droege's and the bead point source (BPS) methods. These MTFs were compared with presampled ones obtained using Boone's method. The results show that Droege's method is not accurate in the low-frequency range, whereas the BPS method is highly sensitive to image noise. While both methods are well adapted to routine stability controls, it was shown that they are not able to provide absolute measurements. On the other hand, Boone's method, which is robust with respect to aliasing, more resilient to noise and provides absolute measurements, satisfies the commissioning requirements perfectly. Thus, Boone's method combined with a modified Catphan 600 phantom could be a good solution to assess CT spatial resolution in the different CT planes.

Degeneracy of antigen recognition as the molecular basis for the high frequency of naive A2/Melan-a peptide multimer(+) CD8(+) T cells in humans.

In contrast with the low frequency of most single epitope reactive T cells in the preimmune repertoire, up to 1 of 1,000 naive CD8(+) T cells from A2(+) individuals specifically bind fluorescent A2/peptide multimers incorporating the A27L analogue of the immunodominant 26-35 peptide from the melanocyte differentiation and melanoma associated antigen Melan-A. This represents the only naive antigen-specific T cell repertoire accessible to direct analysis in humans up to date. To get insight into the molecular basis for the selection and maintenance of such an abundant repertoire, we analyzed the functional diversity of T cells composing this repertoire ex vivo at the clonal level. Surprisingly, we found a significant proportion of multimer(+) clonotypes that failed to recognize both Melan-A analogue and parental peptides in a functional assay but efficiently recognized peptides from proteins of self- or pathogen origin selected for their potential functional cross-reactivity with Melan-A. Consistent with these data, multimers incorporating some of the most frequently recognized peptides specifically stained a proportion of naive CD8(+) T cells similar to that observed with Melan-A multimers. Altogether these results indicate that the high frequency of Melan-A multimer(+) T cells can be explained by the existence of largely cross-reactive subsets of naive CD8(+) T cells displaying multiple specificities.

Automatic quality assessment in structural brain magnetic resonance imaging

MRI has evolved into an important diagnostic technique in medical imaging. However, reliability of the derived diagnosis can be degraded by artifacts, which challenge both radiologists and automatic computer-aided diagnosis. This work proposes a fully-automatic method for measuring image quality of three-dimensional (3D) structural MRI. Quality measures are derived by analyzing the air background of magnitude images and are capable of detecting image degradation from several sources, including bulk motion, residual magnetization from incomplete spoiling, blurring, and ghosting. The method has been validated on 749 3D T(1)-weighted 1.5T and 3T head scans acquired at 36 Alzheimer's Disease Neuroimaging Initiative (ADNI) study sites operating with various software and hardware combinations. Results are compared against qualitative grades assigned by the ADNI quality control center (taken as the reference standard). The derived quality indices are independent of the MRI system used and agree with the reference standard quality ratings with high sensitivity and specificity (>85%). The proposed procedures for quality assessment could be of great value for both research and routine clinical imaging. It could greatly improve workflow through its ability to rule out the need for a repeat scan while the patient is still in the magnet bore.

The CD8 beta polypeptide is required for the recognition of an altered peptide ligand as an agonist.

T cell activation is triggered by the specific recognition of cognate peptides presented by MHC molecules. Altered peptide ligands are analogs of cognate peptides which have a high affinity for MHC molecules. Some of them induce complete T cell responses, i.e. they act as agonists, whereas others behave as partial agonists or even as antagonists. Here, we analyzed both early (intracellular Ca2+ mobilization), and late (interleukin-2 production) signal transduction events induced by a cognate peptide or a corresponding altered peptide ligand using T cell hybridomas expressing or not the CD8 alpha and beta chains. With a video imaging system, we showed that the intracellular Ca2+ response to an altered peptide ligand induces the appearance of a characteristic sustained intracellular Ca2+ concentration gradient which can be detected shortly after T cell interaction with antigen-presenting cells. We also provide evidence that the same altered peptide ligand can be seen either as an agonist or a partial agonist, depending on the presence of CD8beta in the CD8 co-receptor dimers expressed at the T cell surface.

Machine learning for geospatial data : algorithms, software tools and case studies

Résumé Cette thèse est consacrée à l'analyse, la modélisation et la visualisation de données environnementales à référence spatiale à l'aide d'algorithmes d'apprentissage automatique (Machine Learning). L'apprentissage automatique peut être considéré au sens large comme une sous-catégorie de l'intelligence artificielle qui concerne particulièrement le développement de techniques et d'algorithmes permettant à une machine d'apprendre à partir de données. Dans cette thèse, les algorithmes d'apprentissage automatique sont adaptés pour être appliqués à des données environnementales et à la prédiction spatiale. Pourquoi l'apprentissage automatique ? Parce que la majorité des algorithmes d'apprentissage automatiques sont universels, adaptatifs, non-linéaires, robustes et efficaces pour la modélisation. Ils peuvent résoudre des problèmes de classification, de régression et de modélisation de densité de probabilités dans des espaces à haute dimension, composés de variables informatives spatialisées (« géo-features ») en plus des coordonnées géographiques. De plus, ils sont idéaux pour être implémentés en tant qu'outils d'aide à la décision pour des questions environnementales allant de la reconnaissance de pattern à la modélisation et la prédiction en passant par la cartographie automatique. Leur efficacité est comparable au modèles géostatistiques dans l'espace des coordonnées géographiques, mais ils sont indispensables pour des données à hautes dimensions incluant des géo-features. Les algorithmes d'apprentissage automatique les plus importants et les plus populaires sont présentés théoriquement et implémentés sous forme de logiciels pour les sciences environnementales. Les principaux algorithmes décrits sont le Perceptron multicouches (MultiLayer Perceptron, MLP) - l'algorithme le plus connu dans l'intelligence artificielle, le réseau de neurones de régression généralisée (General Regression Neural Networks, GRNN), le réseau de neurones probabiliste (Probabilistic Neural Networks, PNN), les cartes auto-organisées (SelfOrganized Maps, SOM), les modèles à mixture Gaussiennes (Gaussian Mixture Models, GMM), les réseaux à fonctions de base radiales (Radial Basis Functions Networks, RBF) et les réseaux à mixture de densité (Mixture Density Networks, MDN). Cette gamme d'algorithmes permet de couvrir des tâches variées telle que la classification, la régression ou l'estimation de densité de probabilité. L'analyse exploratoire des données (Exploratory Data Analysis, EDA) est le premier pas de toute analyse de données. Dans cette thèse les concepts d'analyse exploratoire de données spatiales (Exploratory Spatial Data Analysis, ESDA) sont traités selon l'approche traditionnelle de la géostatistique avec la variographie expérimentale et selon les principes de l'apprentissage automatique. La variographie expérimentale, qui étudie les relations entre pairs de points, est un outil de base pour l'analyse géostatistique de corrélations spatiales anisotropiques qui permet de détecter la présence de patterns spatiaux descriptible par une statistique. L'approche de l'apprentissage automatique pour l'ESDA est présentée à travers l'application de la méthode des k plus proches voisins qui est très simple et possède d'excellentes qualités d'interprétation et de visualisation. Une part importante de la thèse traite de sujets d'actualité comme la cartographie automatique de données spatiales. Le réseau de neurones de régression généralisée est proposé pour résoudre cette tâche efficacement. Les performances du GRNN sont démontrées par des données de Comparaison d'Interpolation Spatiale (SIC) de 2004 pour lesquelles le GRNN bat significativement toutes les autres méthodes, particulièrement lors de situations d'urgence. La thèse est composée de quatre chapitres : théorie, applications, outils logiciels et des exemples guidés. Une partie importante du travail consiste en une collection de logiciels : Machine Learning Office. Cette collection de logiciels a été développée durant les 15 dernières années et a été utilisée pour l'enseignement de nombreux cours, dont des workshops internationaux en Chine, France, Italie, Irlande et Suisse ainsi que dans des projets de recherche fondamentaux et appliqués. Les cas d'études considérés couvrent un vaste spectre de problèmes géoenvironnementaux réels à basse et haute dimensionnalité, tels que la pollution de l'air, du sol et de l'eau par des produits radioactifs et des métaux lourds, la classification de types de sols et d'unités hydrogéologiques, la cartographie des incertitudes pour l'aide à la décision et l'estimation de risques naturels (glissements de terrain, avalanches). Des outils complémentaires pour l'analyse exploratoire des données et la visualisation ont également été développés en prenant soin de créer une interface conviviale et facile à l'utilisation. Machine Learning for geospatial data: algorithms, software tools and case studies Abstract The thesis is devoted to the analysis, modeling and visualisation of spatial environmental data using machine learning algorithms. In a broad sense machine learning can be considered as a subfield of artificial intelligence. It mainly concerns with the development of techniques and algorithms that allow computers to learn from data. In this thesis machine learning algorithms are adapted to learn from spatial environmental data and to make spatial predictions. Why machine learning? In few words most of machine learning algorithms are universal, adaptive, nonlinear, robust and efficient modeling tools. They can find solutions for the classification, regression, and probability density modeling problems in high-dimensional geo-feature spaces, composed of geographical space and additional relevant spatially referenced features. They are well-suited to be implemented as predictive engines in decision support systems, for the purposes of environmental data mining including pattern recognition, modeling and predictions as well as automatic data mapping. They have competitive efficiency to the geostatistical models in low dimensional geographical spaces but are indispensable in high-dimensional geo-feature spaces. The most important and popular machine learning algorithms and models interesting for geo- and environmental sciences are presented in details: from theoretical description of the concepts to the software implementation. The main algorithms and models considered are the following: multi-layer perceptron (a workhorse of machine learning), general regression neural networks, probabilistic neural networks, self-organising (Kohonen) maps, Gaussian mixture models, radial basis functions networks, mixture density networks. This set of models covers machine learning tasks such as classification, regression, and density estimation. Exploratory data analysis (EDA) is initial and very important part of data analysis. In this thesis the concepts of exploratory spatial data analysis (ESDA) is considered using both traditional geostatistical approach such as_experimental variography and machine learning. Experimental variography is a basic tool for geostatistical analysis of anisotropic spatial correlations which helps to understand the presence of spatial patterns, at least described by two-point statistics. A machine learning approach for ESDA is presented by applying the k-nearest neighbors (k-NN) method which is simple and has very good interpretation and visualization properties. Important part of the thesis deals with a hot topic of nowadays, namely, an automatic mapping of geospatial data. General regression neural networks (GRNN) is proposed as efficient model to solve this task. Performance of the GRNN model is demonstrated on Spatial Interpolation Comparison (SIC) 2004 data where GRNN model significantly outperformed all other approaches, especially in case of emergency conditions. The thesis consists of four chapters and has the following structure: theory, applications, software tools, and how-to-do-it examples. An important part of the work is a collection of software tools - Machine Learning Office. Machine Learning Office tools were developed during last 15 years and was used both for many teaching courses, including international workshops in China, France, Italy, Ireland, Switzerland and for realizing fundamental and applied research projects. Case studies considered cover wide spectrum of the real-life low and high-dimensional geo- and environmental problems, such as air, soil and water pollution by radionuclides and heavy metals, soil types and hydro-geological units classification, decision-oriented mapping with uncertainties, natural hazards (landslides, avalanches) assessments and susceptibility mapping. Complementary tools useful for the exploratory data analysis and visualisation were developed as well. The software is user friendly and easy to use.

Mechanism of hcnA mRNA recognition in the Gac/Rsm signal transduction pathway of Pseudomonas fluorescens.

In the plant-beneficial bacterium Pseudomonas fluorescens CHA0, the expression of antifungal exoproducts is controlled by the GacS/GacA two-component system. Two RNA binding proteins (RsmA, RsmE) ensure effective translational repression of exoproduct mRNAs. At high cell population densities, GacA induces three small RNAs (RsmX, RsmY, RsmZ) which sequester both RsmA and RsmE, thereby relieving translational repression. Here we systematically analyse the features that allow the RNA binding proteins to interact strongly with the 5' untranslated leader mRNA of the P. fluorescens hcnA gene (encoding hydrogen cyanide synthase subunit A). We obtained evidence for three major RsmA/RsmE recognition elements in the hcnA leader, based on directed mutagenesis, RsmE footprints and toeprints, and in vivo expression data. Two recognition elements were found in two stem-loop structures whose existence in the 5' leader region was confirmed by lead(II) cleavage analysis. The third recognition element, which overlapped the hcnA Shine-Dalgarno sequence, was postulated to adopt either an open conformation, which would favour ribosome binding, or a stem-loop structure, which may form upon interaction with RsmA/RsmE and would inhibit access of ribosomes. Effective control of hcnA expression by the Gac/Rsm system appears to result from the combination of the three appropriately spaced recognition elements.

Effects of epitope modification on T cell receptor-ligand binding and antigen recognition by seven H-2Kd-restricted cytotoxic T lymphocyte clones specific for a photoreactive peptide derivative.

We tested for antigen recognition and T cell receptor (TCR)-ligand binding 12 peptide derivative variants on seven H-2Kd-restricted cytotoxic T lymphocytes (CTL) clones specific for a bifunctional photoreactive derivative of the Plasmodium berghei circumsporozoite peptide 252-260 (SYIPSAEKI). The derivative contained iodo-4-azidosalicylic acid in place of PbCS S-252 and 4-azidobenzoic acid on PbCS K-259. Selective photoactivation of the N-terminal photoreactive group allowed crosslinking to Kd molecules and photoactivation of the orthogonal group to TCR. TCR photoaffinity labeling with covalent Kd-peptide derivative complexes allowed direct assessment of TCR-ligand binding on living CTL. In most cases (over 80%) cytotoxicity (chromium release) and TCR-ligand binding differed by less than fivefold. The exceptions included (a) partial TCR agonists (8 cases), for which antigen recognition was five-tenfold less efficient than TCR-ligand binding, (b) TCR antagonists (2 cases), which were not recognized and capable of inhibiting recognition of the wild-type conjugate, (c) heteroclitic agonists (2 cases), for which antigen recognition was more efficient than TCR-ligand binding, and (d) one partial TCR agonist, which activated only Fas (C1)95), but not perforin/granzyme-mediated cytotoxicity. There was no correlation between these divergences and the avidity of TCR-ligand binding, indicating that other factors than binding avidity determine the nature of the CTL response. An unexpected and novel finding was that CD8-dependent clones clearly incline more to TCR antagonism than CD8-independent ones. As there was no correlation between CD8 dependence and the avidity of TCR-ligand binding, the possibility is suggested that CD8 plays a critical role in aberrant CTL function.

Multimodal Strategies in Development.Predictive Value of early simultaneosgesture-speech comination

The present study investigates the predictive value of the early appearance of simultaneous pointing-speech combinations. An experimental task was used to obtain a communicative productive sample from nineteen children at 1;0 and 1;3. Infant’s communicative productions, in combination with gaze joint engagement patterns, were analyzed in relation to different social conditions. The results show a significant effect of age and social condition on infants’ communicative productions. Gesture-speech combinations seem to work as a strong communicative resource to attract the adult’s attention in social demanding communicative contexts. Gaze joint engagement was used in combination with simultaneous pointing-speech combinations to attract adults’ attention during social demanding conditions. Finally, the use of simultaneous pointing-speech combinations at 1;0 in demanding conditions predicted greater expressive vocabulary acquisition at 1;3 and 1;6. These results indicate that the use of gesture-speech combinations may be considered a significant step towards the early integration of language components.