Carregant...
Carregant...

Vés al contingut (premeu Retorn)

Extension of the remos concept to frequency-filtering-based features for reverberation-robust speech recognition

Autor
Maas, R.; Wolf, M.; Sehr, A.; Nadeu, C.; Kellermann, W.
Tipus d'activitat
Presentació treball a congrés
Nom de l'edició
Third Joint Workshop on Hands-free Speech Communication and Microphone Arrays (HSCMA),
Any de l'edició
2011
Data de presentació
2011-06
Llibre d'actes
HSCM A ' 11: 2011 Joint Workshop on Hands-free Speech Communication and Microphone Arrays
Pàgina inicial
13
Pàgina final
18
DOI
https://doi.org/10.1109/HSCMA.2011.5942381 Obrir en finestra nova
Repositori
http://hdl.handle.net/2117/15468 Obrir en finestra nova
URL
http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=5942381&tag=1 Obrir en finestra nova
Resum
The introduction of partly decorrelated features into the REMOS (REverberationMOdeling for Speech recognition) concept for distant-talking speech recognition [1] is discussed. REMOS combines a hidden Markov model (HMM), trained on clean speech, with a reverberation model capturing certain room characteristics. The most likely contributions of both models to a reverberant observation are determined by an inner optimization problem. In HMM frameworks, decorrelated features are assumed when diagona...
Citació
Maas, R. [et al.]. Extension of the remos concept to frequency-filtering-based features for reverberation-robust speech recognition. A: Workshop on Hands-free Speech Communication and Microphone Arrays (HSCMA),. "2011 Joint Workshop on Hands-free Speech Communication and Microphone Arrays". Edinburgh: 2011, p. 13-18.
Grup de recerca
IDEAI-UPC - Intelligent Data Science and Artificial Intelligence Research Center
TALP - Centre de Tecnologies i Aplicacions del Llenguatge i la Parla
VEU - Grup de Tractament de la Parla

Participants

  • Maas, Roland  (autor ponent)
  • Wolf, Martin  (autor ponent)
  • Sehr, Armin  (autor ponent)
  • Nadeu Camprubí, Climent  (autor ponent)
  • Kellermann, Walter  (autor ponent)