Laboratorio de Procesamiento y Transmisión de Voz (LPTV)

Descripción

En el LPTV se realiza investigación y desarrollo en: reconocimiento y verificación del hablante robusto a ruido y canal; CAPT (Computer Aided Pronunciation Tasks); QoS en Internet y procesamiento de señales aplicado a minería, volcanología.

Equipamiento e instrumentos

El LPTV posee una red de PCs, tarjetas DSP y softwares.

Miembros permanentes

Académico responsable

Proyectos asociados

  • PI. "Center for Multidisciplinary Research on Signal Processing", Anillo Ciencia y Tecnología, Conicyt, December 2012-December 2015. US$950.000.
  • Co-PI. "Desarrollo de un sensor continuo de tamaño de burbujas para el proceso de flotación de minerales", CORFO 1IDL2-10687. PI, Prof. Willy Kracht. US$ 50.000.
  • PI, "Robust speech pattern recognition on telephone and education applications". Conicyt/Fondecyt-Chile. From March 2010 to March 2013. US$ 150.000.
  • PI, "Research on adaptation and compensation techniques for speech and speaker recognition". Conicyt/Fondecyt-Chile. From March 2007 to March 2010. US$ 64.000.
  • PI, "ICT technologies for language learning and edutainment in Internet". Conicyt/Fondef-Chile. From March 2007 to December 2009. US$ 265.000.

Publicaciones

  • Néstor Becerra Yoma, Claudio Carretón, Ignacio Catalán, Fernando Huenupan, Jorge Wuth. "On reducing harmonic and sampling distortion in vocal tract length normalization". IEEE Transactions on Audio, Speech and Language Processing, Vol. 21, January, pp. 110-121, 2013.
  • Néstor Becerra Yoma, Leopoldo Benavides, Carlos Molina, Jorge Ruth. "Multicriteria based computer-aided pronunciation quality evaluation of sentences". ETRI Journal, Vol. 35, January, pp.89-99, 2013.
  • Claudio Garretón, Néstor Becerra Yoma. "Telephone channel compensation in text-dependent speaker verification with limited data using a polynomial approximation in the log-filter-bank energy domain". IEEE Transactions on Audio, Speech and Language Processing, Vol. 20, pp.336-341, 2012.
  • Fernando Huenupan, Néstor Becerra Yoma, Claudio Garretón, and Carlos Molina. "Incremental information based on-line optimization of linear combination of classifiers in speaker verification". ETRI (Electronics and Telecommunications Research Institute) journal, Vol. 32, Number 3, June 2010, pages 395-405.
  • Claudio Carretón, Néstor Becerra Yoma, Matías Torres. "Channel robust feature transformation based on filter-bank energy filtering". IEEE Transactions Audio, Speech and Language Processing, Volume 18, Issue 5, pp. 1082-1086, 2010.
  • Juan Pablo Arias, Néstor Becerra Yoma, Hiram Vivanco."Automatic intonation assesment for computer aided language learning". Speech Communciations (Elsevier), Vol. 52, Issue 3, March 2010, pages 254-267.
  • Carlos Molina, Néstor Becerra Yoma, Fernando Huenupán, Claudio Carretón y Jorge Wuth. "Maximum entropy-based reinforcement learning using a confidence measure in speech recognition for telephone speech". In press in IEEE Transactions Audio, Speech and Language Processing, Vol. 18, No. 2, pp.1041-1052, Feb. 2010.
  • Carlos Molina, Néstor Becerra Yoma, Jorge Wuth, Hiram Vivanco. "ASR based pronunciation evaluation with automatically generated competing vocabulary and classifier fusion". Speech Communciations (Elsevier), 51(2009), pag. 485-498, 2009.
  • Néstor Becerra Yoma, Claudio Garretón, Fernando Huenupan y Carlos Molina. "Unsupervised intra-speaker variability compensation based on Gestalt and model adaptation in speaker verification with telephone speech", Speech Communications (Elsevier), Vol. 50/11-12, 953-964, 2008.
  • Fernando Huenupan, Néstor Becerra Yoma, Claudio Garretón, Carlos Molina. Confidence based multiple classifier fusion in speaker verification. Pattern Recognition Letters, Vol 29/7 pp 957-966, 2008.
  • Nestor Becerra Yoma, Carlos Molina. "Feature-dependent compensation of coders in speech recognition". Signal Processing (Elsevier), IEEE Signal Processing Letters, Vol. 86 (1), pp. 38-49, 2006.
  • Alejandro Bassi, Néstor Becerra Yoma, Patricio Loncomilla. "Estimating tonal prosodic discontinuities in Spanish using HMM". Speech Communications (ELsevier), Vol. 48, Pág. 1112-1125, 2006.
  • Ivan Jiron, Ismael Soto, Rolando Carrasco, Néstor Becerra Yoma. "Hyperlliptic curves encryption combined with block codes for Gaussian channel". International journal of communication systems, Volume: 19 Issue: 7 Pages: 809-830, 2006.
Compartir:
https://uchile.cl/i87045
Copiar