Patrol team language identification system for DARPA RATS P1 evaluation

This paper describes the language identification (LID) system developed by the Patrol team for the first phase of the DARPA RATS (Robust Automatic Transcription of Speech) program, which seeks to advance state of the art detection capabilities on audio from highly degraded communication channels. We show that techniques originally developed for LID on telephone speech (e.g., for the NIST language recognition evaluations) remain effective on the noisy RATS data, provided that careful consideration is applied when designing the training and development sets. In addition, we show significant improvements from the use of Wiener filtering, neural network based and language dependent i-vector modeling, and fusion.

Phonotactic language recognition using i-vectors and phoneme posteriogram counts

This paper describes a novel approach to phonotactic LID, where instead of using soft-counts based on phoneme lattices, we use posteriogram to obtain n-gram counts. The high-dimensional vectors of counts are reduced to low-dimensional units for which we adapted the commonly used term i-vectors. The reduction is based on multinomial subspace modeling and is designed to work in the total-variability space. The proposed technique was tested on the NIST 2009 LRE set with better results to a system based on using soft-counts (Cavg on 30s: 3.15% vs 3.43%), and with very good results when fused with an acoustic i-vector LID system (Cavg on 30s acoustic 2.4% vs 1.25%). The proposed technique is also compared with another low dimensional projection system based on PCA. In comparison with the original soft-counts, the proposed technique provides better results, reduces the problems due to sparse counts, and avoids the process of using pruning techniques when creating the lattices.

Low-resource language recognition using a fusion of phoneme posteriorgram counts, acoustic and glottal-based i-vectors

This paper presents a description of our system for the Albayzin 2012 LRE competition. One of the main characteristics of this evaluation was the reduced number of available files for training the system, especially for the empty condition where no training data set was provided but only a development set. In addition, the whole database was created from online videos and around one third of the training data was labeled as noisy files. Our primary system was the fusion of three different i-vector based systems: one acoustic system based on MFCCs, a phonotactic system using trigrams of phone-posteriorgram counts, and another acoustic system based on RPLPs that improved robustness against noise. A contrastive system that included new features based on the glottal source was also presented. Official and postevaluation results for all the conditions using the proposed metrics for the evaluation and the Cavg metric are presented in the paper.

Adapting a speech into sign language translation system to a new domain

This paper presents a methodology for adapting an advanced communication system for deaf people in a new domain. This methodology is a user-centered design approach consisting of four main steps: requirement analysis, parallel corpus generation, technology adaptation to the new domain, and finally, system evaluation. In this paper, the new considered domain has been the dialogues in a hotel reception. With this methodology, it was possible to develop the system in a few months, obtaining very good performance: good speech recognition and translation rates (around 90%) with small processing times.

Smart aeronautical structures: development and experimental validation of a structural health monitoring system for damage detection

En muchas áreas de la ingeniería, la integridad y confiabilidad de las estructuras son aspectos de extrema importancia. Estos son controlados mediante el adecuado conocimiento de danos existentes. Típicamente, alcanzar el nivel de conocimiento necesario que permita caracterizar la integridad estructural implica el uso de técnicas de ensayos no destructivos. Estas técnicas son a menudo costosas y consumen mucho tiempo. En la actualidad, muchas industrias buscan incrementar la confiabilidad de las estructuras que emplean. Mediante el uso de técnicas de última tecnología es posible monitorizar las estructuras y en algunos casos, es factible detectar daños incipientes que pueden desencadenar en fallos catastróficos. Desafortunadamente, a medida que la complejidad de las estructuras, los componentes y sistemas incrementa, el riesgo de la aparición de daños y fallas también incrementa. Al mismo tiempo, la detección de dichas fallas y defectos se torna más compleja. En años recientes, la industria aeroespacial ha realizado grandes esfuerzos para integrar los sensores dentro de las estructuras, además de desarrollar algoritmos que permitan determinar la integridad estructural en tiempo real. Esta filosofía ha sido llamada “Structural Health Monitoring” (o “Monitorización de Salud Estructural” en español) y este tipo de estructuras han recibido el nombre de “Smart Structures” (o “Estructuras Inteligentes” en español). Este nuevo tipo de estructuras integran materiales, sensores, actuadores y algoritmos para detectar, cuantificar y localizar daños dentro de ellas mismas. Una novedosa metodología para detección de daños en estructuras se propone en este trabajo. La metodología está basada en mediciones de deformación y consiste en desarrollar técnicas de reconocimiento de patrones en el campo de deformaciones. Estas últimas, basadas en PCA (Análisis de Componentes Principales) y otras técnicas de reducción dimensional. Se propone el uso de Redes de difracción de Bragg y medidas distribuidas como sensores de deformación. La metodología se validó mediante pruebas a escala de laboratorio y pruebas a escala real con estructuras complejas. Los efectos de las condiciones de carga variables fueron estudiados y diversos experimentos fueron realizados para condiciones de carga estáticas y dinámicas, demostrando que la metodología es robusta ante condiciones de carga desconocidas. ABSTRACT In many engineering fields, the integrity and reliability of the structures are extremely important aspects. They are controlled by the adequate knowledge of existing damages. Typically, achieving the level of knowledge necessary to characterize the structural integrity involves the usage of nondestructive testing techniques. These are often expensive and time consuming. Nowadays, many industries look to increase the reliability of the structures used. By using leading edge techniques it is possible to monitoring these structures and in some cases, detect incipient damage that could trigger catastrophic failures. Unfortunately, as the complexity of the structures, components and systems increases, the risk of damages and failures also increases. At the same time, the detection of such failures and defects becomes more difficult. In recent years, the aerospace industry has done great efforts to integrate the sensors within the structures and, to develop algorithms for determining the structural integrity in real time. The ‘philosophy’ has being called “Structural Health Monitoring” and these structures have been called “smart structures”. These new types of structures integrate materials, sensors, actuators and algorithms to detect, quantify and locate damage within itself. A novel methodology for damage detection in structures is proposed. The methodology is based on strain measurements and consists in the development of strain field pattern recognition techniques. The aforementioned are based on PCA (Principal Component Analysis) and other dimensional reduction techniques. The use of fiber Bragg gratings and distributed sensing as strain sensors is proposed. The methodology have been validated by using laboratory scale tests and real scale tests with complex structures. The effects of the variable load conditions were studied and several experiments were performed for static and dynamic load conditions, demonstrating that the methodology is robust under unknown load conditions.

Analysis of pattern recognition and dimensionality reduction techniques for odor biometrics

In this paper, we analyze the performance of several well-known pattern recognition and dimensionality reduction techniques when applied to mass-spectrometry data for odor biometric identification. Motivated by the successful results of previous works capturing the odor from other parts of the body, this work attempts to evaluate the feasibility of identifying people by the odor emanated from the hands. By formulating this task according to a machine learning scheme, the problem is identified with a small-sample-size supervised classification problem in which the input data is formed by mass spectrograms from the hand odor of 13 subjects captured in different sessions. The high dimensionality of the data makes it necessary to apply feature selection and extraction techniques together with a simple classifier in order to improve the generalization capabilities of the model. Our experimental results achieve recognition rates over 85% which reveals that there exists discriminatory information in the hand odor and points at body odor as a promising biometric identifier.

Speech Signals Feature Extraction Model for a Speaker’s Gender and Age Identification System

Durante el proceso de producción de voz, los factores anatómicos, fisiológicos o psicosociales del individuo modifican los órganos resonadores, imprimiendo en la voz características particulares. Los sistemas ASR tratan de encontrar los matices característicos de una voz y asociarlos a un individuo o grupo. La edad y sexo de un hablante son factores intrínsecos que están presentes en la voz. Este trabajo intenta diferenciar esas características, aislarlas y usarlas para detectar el género y la edad de un hablante. Para dicho fin, se ha realizado el estudio y análisis de las características basadas en el pulso glótico y el tracto vocal, evitando usar técnicas clásicas (como pitch y sus derivados) debido a las restricciones propias de dichas técnicas. Los resultados finales de nuestro estudio alcanzan casi un 100% en reconocimiento de género mientras en la tarea de reconocimiento de edad el reconocimiento se encuentra alrededor del 80%. Parece ser que la voz queda afectada por el género del hablante y las hormonas, aunque no se aprecie en la audición. ABSTRACT Particular elements of the voice are printed during the speech production process and are related to anatomical and physiological factors of the phonatory system or psychosocial factors acquired by the speaker. ASR systems attempt to find those peculiar nuances of a voice and associate them to an individual or a group. Age and gender are inherent factors to the speaker which may be represented in voice. This work attempts to differentiate those characteristics, isolate them and use them to detect speaker’s gender and age. Features based on glottal pulse and vocal tract are studied and analyzed in order to achieve good results in both tasks. Classical methodologies (such as pitch and derivates) are avoided since the requirements of those techniques may be too restrictive. The final scores achieve almost 100% in gender recognition whereas in age recognition those scores are around 80%. Factors related to the gender and hormones seem to affect the voice although they are not audible.

Extended phone log-likelihood ratio features and acoustic-based I-vectors for language recognition

This paper presents new techniques with relevant improvements added to the primary system presented by our group to the Albayzin 2012 LRE competition, where the use of any additional corpora for training or optimizing the models was forbidden. In this work, we present the incorporation of an additional phonotactic subsystem based on the use of phone log-likelihood ratio features (PLLR) extracted from different phonotactic recognizers that contributes to improve the accuracy of the system in a 21.4% in terms of Cavg (we also present results for the official metric during the evaluation, Fact). We will present how using these features at the phone state level provides significant improvements, when used together with dimensionality reduction techniques, especially PCA. We have also experimented with applying alternative SDC-like configurations on these PLLR features with additional improvements. Also, we will describe some modifications to the MFCC-based acoustic i-vector system which have also contributed to additional improvements. The final fused system outperformed the baseline in 27.4% in Cavg.

Language recognition using phonotactic-based shifted delta coefficients and multiple phone recognizers

A new language recognition technique based on the application of the philosophy of the Shifted Delta Coefficients (SDC) to phone log-likelihood ratio features (PLLR) is described. The new methodology allows the incorporation of long-span phonetic information at a frame-by-frame level while dealing with the temporal length of each phone unit. The proposed features are used to train an i-vector based system and tested on the Albayzin LRE 2012 dataset. The results show a relative improvement of 33.3% in Cavg in comparison with different state-of-the-art acoustic i-vector based systems. On the other hand, the integration of parallel phone ASR systems where each one is used to generate multiple PLLR coefficients which are stacked together and then projected into a reduced dimension are also presented. Finally, the paper shows how the incorporation of state information from the phone ASR contributes to provide additional improvements and how the fusion with the other acoustic and phonotactic systems provides an important improvement of 25.8% over the system presented during the competition.

Automatic physical activity pattern recognition and meal intake modelling in type 1 diabetes for their inclusion in artificial pancreas systems

La diabetes comprende un conjunto de enfermedades metabólicas que se caracterizan por concentraciones de glucosa en sangre anormalmente altas. En el caso de la diabetes tipo 1 (T1D, por sus siglas en inglés), esta situación es debida a una ausencia total de secreción endógena de insulina, lo que impide a la mayoría de tejidos usar la glucosa. En tales circunstancias, se hace necesario el suministro exógeno de insulina para preservar la vida del paciente; no obstante, siempre con la precaución de evitar caídas agudas de la glucemia por debajo de los niveles recomendados de seguridad. Además de la administración de insulina, las ingestas y la actividad física son factores fundamentales que influyen en la homeostasis de la glucosa. En consecuencia, una gestión apropiada de la T1D debería incorporar estos dos fenómenos fisiológicos, en base a una identificación y un modelado apropiado de los mismos y de sus sorrespondientes efectos en el balance glucosa-insulina. En particular, los sistemas de páncreas artificial –ideados para llevar a cabo un control automático de los niveles de glucemia del paciente– podrían beneficiarse de la integración de esta clase de información. La primera parte de esta tesis doctoral cubre la caracterización del efecto agudo de la actividad física en los perfiles de glucosa. Con este objetivo se ha llevado a cabo una revisión sistemática de la literatura y meta-análisis que determinen las respuestas ante varias modalidades de ejercicio para pacientes con T1D, abordando esta caracterización mediante unas magnitudes que cuantifican las tasas de cambio en la glucemia a lo largo del tiempo. Por otro lado, una identificación fiable de los periodos con actividad física es un requisito imprescindible para poder proveer de esa información a los sistemas de páncreas artificial en condiciones libres y ambulatorias. Por esta razón, la segunda parte de esta tesis está enfocada a la propuesta y evaluación de un sistema automático diseñado para reconocer periodos de actividad física, clasificando su nivel de intensidad (ligera, moderada o vigorosa); así como, en el caso de periodos vigorosos, identificando también la modalidad de ejercicio (aeróbica, mixta o de fuerza). En este sentido, ambos aspectos tienen una influencia específica en el mecanismo metabólico que suministra la energía para llevar a cabo el ejercicio y, por tanto, en las respuestas glucémicas en T1D. En este trabajo se aplican varias combinaciones de técnicas de aprendizaje máquina y reconocimiento de patrones sobre la fusión multimodal de señales de acelerometría y ritmo cardíaco, las cuales describen tanto aspectos mecánicos del movimiento como la respuesta fisiológica del sistema cardiovascular ante el ejercicio. Después del reconocimiento de patrones se incorpora también un módulo de filtrado temporal para sacar partido a la considerable coherencia temporal presente en los datos, una redundancia que se origina en el hecho de que en la práctica, las tendencias en cuanto a actividad física suelen mantenerse estables a lo largo de cierto tiempo, sin fluctuaciones rápidas y repetitivas. El tercer bloque de esta tesis doctoral aborda el tema de las ingestas en el ámbito de la T1D. En concreto, se propone una serie de modelos compartimentales y se evalúan éstos en función de su capacidad para describir matemáticamente el efecto remoto de las concetraciones plasmáticas de insulina exógena sobre las tasas de eleiminación de la glucosa atribuible a la ingesta; un aspecto hasta ahora no incorporado en los principales modelos de paciente para T1D existentes en la literatura. Los datos aquí utilizados se obtuvieron gracias a un experimento realizado por el Institute of Metabolic Science (Universidad de Cambridge, Reino Unido) con 16 pacientes jóvenes. En el experimento, de tipo ‘clamp’ con objetivo variable, se replicaron los perfiles individuales de glucosa, según lo observado durante una visita preliminar tras la ingesta de una cena con o bien alta carga glucémica, o bien baja. Los seis modelos mecanísticos evaluados constaban de: a) submodelos de doble compartimento para las masas de trazadores de glucosa, b) un submodelo de único compartimento para reflejar el efecto remoto de la insulina, c) dos tipos de activación de este mismo efecto remoto (bien lineal, bien con un punto de corte), y d) diversas condiciones iniciales. ABSTRACT Diabetes encompasses a series of metabolic diseases characterized by abnormally high blood glucose concentrations. In the case of type 1 diabetes (T1D), this situation is caused by a total absence of endogenous insulin secretion, which impedes the use of glucose by most tissues. In these circumstances, exogenous insulin supplies are necessary to maintain patient’s life; although caution is always needed to avoid acute decays in glycaemia below safe levels. In addition to insulin administrations, meal intakes and physical activity are fundamental factors influencing glucose homoeostasis. Consequently, a successful management of T1D should incorporate these two physiological phenomena, based on an appropriate identification and modelling of these events and their corresponding effect on the glucose-insulin balance. In particular, artificial pancreas systems –designed to perform an automated control of patient’s glycaemia levels– may benefit from the integration of this type of information. The first part of this PhD thesis covers the characterization of the acute effect of physical activity on glucose profiles. With this aim, a systematic review of literature and metaanalyses are conduced to determine responses to various exercise modalities in patients with T1D, assessed via rates-of-change magnitudes to quantify temporal variations in glycaemia. On the other hand, a reliable identification of physical activity periods is an essential prerequisite to feed artificial pancreas systems with information concerning exercise in ambulatory, free-living conditions. For this reason, the second part of this thesis focuses on the proposal and evaluation of an automatic system devised to recognize physical activity, classifying its intensity level (light, moderate or vigorous) and for vigorous periods, identifying also its exercise modality (aerobic, mixed or resistance); since both aspects have a distinctive influence on the predominant metabolic pathway involved in fuelling exercise, and therefore, in the glycaemic responses in T1D. Various combinations of machine learning and pattern recognition techniques are applied on the fusion of multi-modal signal sources, namely: accelerometry and heart rate measurements, which describe both mechanical aspects of movement and the physiological response of the cardiovascular system to exercise. An additional temporal filtering module is incorporated after recognition in order to exploit the considerable temporal coherence (i.e. redundancy) present in data, which stems from the fact that in practice, physical activity trends are often maintained stable along time, instead of fluctuating rapid and repeatedly. The third block of this PhD thesis addresses meal intakes in the context of T1D. In particular, a number of compartmental models are proposed and compared in terms of their ability to describe mathematically the remote effect of exogenous plasma insulin concentrations on the disposal rates of meal-attributable glucose, an aspect which had not yet been incorporated to the prevailing T1D patient models in literature. Data were acquired in an experiment conduced at the Institute of Metabolic Science (University of Cambridge, UK) on 16 young patients. A variable-target glucose clamp replicated their individual glucose profiles, observed during a preliminary visit after ingesting either a high glycaemic-load or a low glycaemic-load evening meal. The six mechanistic models under evaluation here comprised: a) two-compartmental submodels for glucose tracer masses, b) a single-compartmental submodel for insulin’s remote effect, c) two types of activations for this remote effect (either linear or with a ‘cut-off’ point), and d) diverse forms of initial conditions.

Immune Polarization in Allergic Patients: Role of the Innate Immune System

Allergens come into contact with the immune system as components of a very diverse mixture. The most common sources are pollen grains, food, and waste. These sources contain a variety of immunomodulatory components that play a key role in the induction of allergic sensitization. The way allergen molecules bind to the cells of the immune system can determine the immune response. In order to better understand how allergic sensitization is triggered, we review the molecular mechanisms involved in the development of allergy and the role of immunomodulators in allergen recognition by innate cells.

Polymerase recognition of synthetic oligodeoxyribonucleotides incorporating degenerate pyrimidine and purine bases

A universal base that is capable of substituting for any of the four natural bases in DNA would be of great utility in both mutagenesis and recombinant DNA experiments. This paper describes the properties of oligonucleotides incorporating two degenerate bases, the pyrimidine base 6H,8H-3,4-dihydropyrimido[4,5-c][1,2]oxazin-7-one and the purine base N6-methoxy-2,6-diaminopurine, designated P and K, respectively. An equimolar mixture of the analogues P and K (called M) acts, in primers, as a universal base. The thermal stability of oligonucleotide duplexes were only slightly reduced when natural bases were replaced by P or K. Templates containing the modified bases were copied by Taq polymerase; P behaved as thymine in 60% of copying events and as cytosine in 40%, whereas K behaved as if it were guanine (13%) or adenine (87%). The dUTPase gene of Caenorhabditis elegans, which we have found to contain three nonidentical homologous repeats, was used as a model system to test the use of these bases in primers for DNA synthesis. A pair of oligodeoxyribonucleotides, each 20 residues long and containing an equimolar mixture of P and K at six positions, primed with high specificity both T7 DNA polymerase in sequencing reactions and Taq polymerase in PCRs; no nonspecific amplification was obtained on genomic DNA of C. elegans. Use of P and K can significantly reduce the complexity of degenerate oligonucleotide mixtures, and when used together, P and K can act as a universal base.

Diversification, expression, and γδ T cell recognition of evolutionarily distant members of the MIC family of major histocompatibility complex class I-related molecules

Distant relatives of major histocompatibility complex (MHC) class I molecules, human MICA and MICB, function as stress-induced antigens that are broadly recognized by intestinal epithelial γδ T cells. They may thus play a central role in the immune surveillance of damaged, infected, or otherwise stressed intestinal epithelial cells. However, the generality of this system in evolution and the mode of recognition of MICA and MICB are undefined. Analysis of cDNA sequences from various primate species defined translation products that are homologous to MICA and MICB. All of the MIC polypeptides have common characteristics, although they are extraordinarily diverse. The most notable alterations are several deletions and frequent amino acid substitutions in the putative α-helical regions of the α1α2 domains. However, the primate MIC molecules were expressed on the surfaces of normal and transfected cells. Moreover, despite their sharing of relatively few identical amino acids in potentially accessible regions of their α1α2 domains, they were recognized by diverse human intestinal epithelial γδ T cells that are restricted by MICA and MICB. Thus, MIC molecules represent a family of MHC proteins that are structurally diverse yet appear to be functionally conserved. The promiscuous mode of γδ T cell recognition of these antigens may be explained by their sharing of a single conserved interaction site.

Identification and reconstitution of the origin recognition complex from Schizosaccharomyces pombe

The origin recognition complex (ORC), first identified in Saccharomyces cerevisiae (sc), is a six-subunit protein complex that binds to DNA origins. Here, we report the identification and cloning of cDNAs encoding the six subunits of the ORC of Schizosaccharomyces pombe (sp). Sequence analyses revealed that spOrc1, 2, and 5 subunits are highly conserved compared with their counterparts from S. cerevisiae, Xenopus, Drosophila, and human. In contrast, both spOrc3 and spOrc6 subunits are poorly conserved. As reported by Chuang and Kelly [(1999) Proc. Natl. Acad. Sci. USA 96, 2656–2661], the C-terminal region of spOrc4 is also conserved whereas the N terminus uniquely contains repeats of a sequence that binds strongly to AT-rich DNA regions. Consistent with this, extraction of S. pombe chromatin with 1 M NaCl, or after DNase I treatment, yielded the six-subunit ORC, whereas extraction with 0.3 M resulted in five-subunit ORC lacking spOrc4p. The spORC can be reconstituted in vitro with all six recombinant subunits expressed in the rabbit reticulocyte system. The association of spOrc4p with the other subunits required the removal of DNA from reaction mixture by DNase I. This suggests that a strong interaction between spOrc4p and DNA can prevent the isolation of the six-subunit ORC. The unique DNA-binding properties of the spORC may contribute to our understanding of the sequence-specific recognition required for the initiation of DNA replication in S. pombe.

Specific atomic groups and RNA helix geometry in acceptor stem recognition by a tRNA synthetase

Oligonucleotides that recapitulate the acceptor stems of tRNAs are substrates for aminoacylation by many tRNA synthetases in vitro, even though these substrates are missing the anticodon trinucleotides of the genetic code. In the case of tRNAAla a single acceptor stem G⋅U base pair at position 3·70 is essential, based on experiments where the wobble pair has been replaced by alternatives such as I⋅U, G⋅C, and A⋅U, among others. These experiments led to the conclusion that the minor-groove free 2-amino group (of guanosine) of the G⋅U wobble pair is essential for charging. Moreover, alanine-inserting tRNAs (amber suppressors) that replace G⋅U with mismatches such as G⋅A and C⋅A are partially active in vivo and can support growth of an Escherichia coli tRNAAla knockout strain, leading to the hypothesis that a helix irregularity and nucleotide functionalities are important for recognition. Herein we investigate the charging in vitro of oligonucleotide and full-length tRNA substrates that contain mismatches at the position of the G⋅U pair. Although most of these substrates have undetectable activity, G⋅A and C⋅A variants retain some activity, which is, nevertheless, reduced by at least 100-fold. Thus, the in vivo assays are much less sensitive to large changes in aminoacylation kinetic efficiency of 3·70 variants than is the in vitro assay system. Although these functional data do not clarify all of the details, it is now clear that specific atomic groups are substantially more important in determining kinetic efficiency than is a helical distortion. By implication, the activity of mutant tRNAs measured in the in vivo assays appears to be more dependent on factors other than aminoacylation kinetic efficiency.