THALES-UPM

 

  • José Manuel Pardo (PI)
  • Ricardo de Córdoba
  • Javier Ferreiros
  • José Carlos González
  • José Miguel Goñi
  • Juan Manuel Montero
  • Verónica López
  • Julián David Echeberry
  • Beatriz Martínez González
  • Juan Fernández Fernández
  • Sara Lana
  • Julio Villena

The group Tecnologías de Audio, Habla y Lenguaje Natural en Sistemas Inteligentes is formed by researchers from the Grupo de Sistemas Inteligentes and Grupo de Tecnología del Habla at Universidad Politécnica de Madrid who have joined forces to create a multidisciplinary team with the aim of working on one of the basic objectives of MAVIR: the access and retrieval of multilingual and multimedia information.

The members of Grupo de Sistemas Inteligentes are working on Human Language Technology since 1990. They collaborate with the Laboratorio de Lingüística Informática from the Universidad Autónoma de Madrid and with the Laboratorio de Bases de Datos Avanzadas from the Universidad Carlos III de Madrid. The research areas include the development of linguistic resources in Spanish, algorithms and data structures for an efficient management of the resources, and the development of tools for part of speech analysis. The group has registered resources from the ARIES platform and library for structured management from TRIE in the Registro de la Propiedad Intelectual (Intellectual Property Registry). This technology has been transferred to industry by DAEDALUS, S.A., a spin-off of the group founded in 1998). With respect to NLP applications, the Grupo de Sistemas Inteligentes has worked on multilingual information retrieval evaluation tasks organised by CLEF.

The group Tecnología del Habla is working since the 1970s on automatic speech recognition, identification and verification of speakers, automatic speech segmentation by speaker, language recognition, natural language understanding, conversion of emotional speech to text, human-machine dialogue systems in intelligent environments, automatic machine translation, with wide experience on voice translation to LSE tools for users with disabilities.

The group has been developing since 1978 high quality voice synthesis systems in Spanish, and of speech recognition, dialogue and understanding in various vocabulary, speakers and type of speech conditions. The group has collaborated in technology transfer with various national and international companies and has commercialised produced with companies such as Speech Plus, Digital Equipment Corporation, Telia Promotor Infovox, Hewlett Packard, Dygra etc., and has integrated tools and systems in industrial environments. It has participated in regional, national and European funded projects (ESPRIT, LRE, LE, TIDE, IST), showing excellence in systems and tools related to speech technology.

In 1993 the group was awarded the Premio Reina Sofía 1993 de Rehabilitación e Integración and in 1995 organised the international conference EUROSPEECH 95 in Madrid. It has also participated in the NATO IST 03 group and in the European Language and Speech Network Executive Board. The group's researchers have visited prestigious international research centres, such as MIT, UC Berkeley, University of Colorado, Cambridge University, University of Aachen, University of Edinburgh and SRI. A group member has participated in the diarisation international evaluation task on speech retrieval (with ICSI at Berkeley) obtaining the best scoring results, and will participate again this year.

The following projects in which the group has been involved during the last five years, are directly related to the current proposal: ISAIAS: Desarrollo de Tecnología del habla para el tratamiento automático de información en reuniones y seminarios (TIC2000-0198-P4-04, 2001-2002); INVOCA: Interfaces vocales para control de tráfico aéreo (funded by AENA, 2001-2003); GEMINI: Generic Environment for Multilingual Interactive Natural Interfaces (EU-IST-2001-32343, 2002-2004); RESINTRA: Apoyo a la introducción de las técnicas de reconocimiento y síntesis de voz en los procesos de simulación y cálculo de carga de trabajo (funded by AENA, 2005-2008); Study on the Prospective of Speech-based Interfaces for future Digital Services (150794-2007 F1ED-ES, EU, 2007-2008); Desarrollo de un sistema de traducción de voz a lengua de signos española para un servicio público de atención personal (Plan Avanza, PAV-070000-2007-567, 2007-2008); ROBONAUTA: Integración de sistemas de reconocimiento en entornos de robótica móvil, reconocimiento automático de habla en entornos hostiles y en mecanismos de respuesta basados en síntesis de voz con emociones (MCYT DPI2007-66846-c02-02, 2007-2010); SD-TEAM: Modelos dinámicos de gestión de diálogo en sistemas multimodales y multidominio (MCYT, TIN2008-06856-C05-03/TIN, 2008- 2011); BRAVO: Búsqueda de Respuestas Avanzada Multimodal y Multilingüe (TIN2007-67407-C03-03, 2007-2010); ITECBAN: Infraestructura Tecnológica y Metodológica de Soporte para un Core Bancario (Programa CENIT, 2006-2010); RIMMEL: Recuperación de Información Multilingüe y Multimedia y su Evaluación (MCYT, TIN2004-07588-C03-01, 2004-2007); SADE: Servicios Avanzados Domóticos e Inmóticos en MIDAS (MCYT, TIC2003-09192-C11-06, 2003-2004); MIRACLE: Sistema de Recuperación de Información Multilingüe y su Evaluación por la Iniciativa Europea CLEF (Comunidad de Madrid TIC 07T/0055, 2003-2004); and OMNIPAPER: Smart Access to European Newspapers (ESPRIT. IST-2001-32174, 2002-2004).

 
RocketTheme Joomla Templates