logo Idiap Research Institute        
 [BibTeX] [Marc21]
Grapheme-based Automatic Speech Recognition using KL-HMM
Type of publication: Conference paper
Citation: Magimai.-Doss_INTERSPEECH2011_2011
Publication status: Accepted
Booktitle: Proceedings of Interspeech
Year: 2011
Month: August
Abstract: The state-of-the-art automatic speech recognition (ASR) systems typically use phonemes as subword units. In this work, we present a novel grapheme-based ASR system that jointly models phoneme and grapheme information using Kullback-Leibler divergence-based HMM system (KL-HMM). More specifically, the underlying subword unit models are grapheme units and the phonetic information is captured through phoneme posterior probabilities (referred as posterior features) estimated using a multilayer perceptron (MLP). We investigate the proposed approach for ASR on English language, where the correspondence between phoneme and grapheme is weak. In particular, we investigate the effect of contextual modeling on grapheme-based KL-HMM system and the use of MLP trained on auxiliary data. Experiments on DARPA Resource Management corpus have shown that the grapheme-based ASR system modeling longer subword unit context can achieve same performance as phoneme-based ASR system, irrespective of the data on which MLP is trained.
Keywords: Automatic Speech Recognition, Graphemes, Kullback-Leibler divergence based hidden Markov model, multilayer perceptron, phonemes, Posterior features
Projects Idiap
Authors Magimai.-Doss, Mathew
Rasipuram, Ramya
Aradilla, Guillermo
Bourlard, Hervé
Added by: [UNK]
Total mark: 0
  • Magimai.-Doss_INTERSPEECH2011_2011.pdf