121 search results - page 11 / 25 » Corrected tandem features for acoustic model training |
ICASSP
2010
IEEE
13 years 7 months ago
2010
IEEE
This paper describes the Arabic broadcast transcription system fielded by IBM in the GALE Phase 3.5 machine translation evaluation. Key advances compared to our Phase 2.5 system ...
ICASSP
2011
IEEE
12 years 11 months ago
2011
IEEE
A novel frame-wise model adaptation approach for reverberationrobust distant-talking speech recognition is proposed. It adjusts the means of static cepstral features to capture th...
JMLR
2010
13 years 2 months ago
2010
L1 (also referred to as the 1-norm or Lasso) penalty based formulations have been shown to be effective in problem domains when noisy features are present. However, the L1 penalty...
INTERSPEECH
2010
13 years 2 months ago
2010
In this paper we evaluate a method for generating synthetic speech at high speaking rates based on the interpolation of hidden semi-Markov models (HSMMs) trained on speech data re...
NAACL
1994
13 years 9 months ago
1994
Improved acoustic modeling can significantly decrease the error rate in large-vocabulary speech recognition. Our approach to the problem is twofold. We first propose a scheme that...