Deep-neural network approaches for speech recognition with heterogeneous groups of speakers including children
- 12 April 2016
- journal article
- research article
- Published by Cambridge University Press (CUP) in Natural Language Engineering
- Vol. 23 (3) , 325-350
Abstract
This paper introduces deep neural network (DNN)–hidden Markov model (HMM)-based methods to tackle speech recognition in heterogeneous groups of speakers including children. We target three speaker groups consisting of children, adult males and adult females. Two different kind of approaches are introduced here: approaches based on DNN adaptation and approaches relying on vocal-tract length normalisation (VTLN). First, the recent approach that consists in adapting a general DNN to domain/language specific data is extended to target age/gender groups in the context of DNN–HMM. Then, VTLN is investigated by training a DNN–HMM system by using either mel frequency cepstral coefficients normalised with standard VTLN or mel frequency cepstral coefficients derived acoustic features combined with the posterior probabilities of the VTLN warping factors. In this later, novel, approach the posterior probabilities of the warping factors are obtained with a separate DNN and the decoding can be operated in a single pass when the VTLN approach requires two decoding passes. Finally, the different approaches presented here are combined to take advantage of their complementarity. The combination of several approaches is shown to improve the baseline phone error rate performance by thirty per cent to thirty-five per cent relative and the baseline word error rate performance by about ten per cent relative.Keywords
This publication has 34 references indexed in Scilit:
- Representation Learning: A Review and New PerspectivesIEEE Transactions on Pattern Analysis and Machine Intelligence, 2013
- Tandem decoding of children's speech for keyword detection in a child-robot interaction scenarioACM Transactions on Speech and Language Processing, 2011
- Parallel Training of Neural Networks for Speech RecognitionLecture Notes in Computer Science, 2010
- Towards age-independent acoustic modelingSpeech Communication, 2009
- Acoustic variability and automatic recognition of children’s speechSpeech Communication, 2007
- A Fast Learning Algorithm for Deep Belief NetsNeural Computation, 2006
- Robust recognition of children's speechIEEE Transactions on Speech and Audio Processing, 2003
- Heteroscedastic discriminant analysis and reduced rank HMMs for improved speech recognitionSpeech Communication, 1998
- Maximum likelihood linear transformations for HMM-based speech recognitionComputer Speech & Language, 1998
- A novel feature transformation for vocal tract length normalization in automatic speech recognitionIEEE Transactions on Speech and Audio Processing, 1998