Unsupervised discriminative adaptation using discriminative mapping transforms
Authors:
- 1 March 2008
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE) in International Conference on Acoustics, Speech, and Signal Processing (ICASSP)
- No. 15206149,p. 4273-4276
Abstract
The most commonly used approaches to speaker adaptation are based on linear transforms, as these can be robustly estimated using limited adaptation data. Although significant gains can be obtained using discriminative criteria for training acoustic models, maximum likelihood (ML) estimated transforms are used for unsupervised adaptation. This is because discriminatively trained transforms are highly sensitive to errors in the adaptation hypothesis. This paper describes a new framework for estimating transforms that are discriminative in nature, but are less sensitive to this hypothesis issue. A discriminative, speaker-independent, mapping transformation is estimated during training. This transform is obtained after a speaker-specific ML-estimated transform has been applied. During recognition an ML speaker-specific transform is found and the speaker-independent discriminative mapping transform then applied. This allows a transform which is discriminative in nature to be indirectly estimated, whilst only requiring an ML speaker-specific transform to be found during recognition. The scheme is evaluated on an English conversational telephone speech task, where it significantly outperforms both standard ML and discriminatively trained transforms.Keywords
This publication has 7 references indexed in Scilit:
- Discriminative cluster adaptive trainingIEEE Transactions on Audio, Speech, and Language Processing, 2006
- Discriminative linear transforms for feature normalization and speaker adaptation in HMM estimationIEEE Transactions on Speech and Audio Processing, 2005
- Discriminative adaptive training using the MPE criterion2004
- Minimum phone error and I-smoothing for improved discriminative training2002
- Maximum likelihood linear transformations for HMM-based speech recognitionComputer Speech & Language, 1998
- Maximum likelihood linear regression for speaker adaptation of continuous density hidden Markov modelsComputer Speech & Language, 1995
- Speaker adaptation using constrained estimation of Gaussian mixturesIEEE Transactions on Speech and Audio Processing, 1995