iVector-based discriminative adaptation for automatic speech recognition
- 1 December 2011
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 152-157
Abstract
We presented a novel technique for discriminative feature-level adaptation of automatic speech recognition system. The concept of iVectors popular in Speaker Recognition is used to extract information about speaker or acoustic environment from speech segment. iVector is a low-dimensional fixed-length representing such information. To utilized iVectors for adaptation, Region Dependent Linear Transforms (RDLT) are discriminatively trained using MPE criterion on large amount of annotated data to extract the relevant information from iVectors and to compensate speech feature. The approach was tested on standard CTS data. We found it to be complementary to common adaptation techniques. On a well tuned RDLT system with standard CMLLR adaptation we reached 0.8% additive absolute WER improvement.Keywords
This publication has 10 references indexed in Scilit:
- Full-covariance UBM and heavy-tailed PLDA in i-vector speaker verification2011
- Simplification and optimization of i-vector extraction2011
- The subspace Gaussian mixture model—A structured model for speech recognitionComputer Speech & Language, 2011
- Front-End Factor Analysis for Speaker VerificationIEEE Transactions on Audio, Speech, and Language Processing, 2010
- Unsupervised discriminative adaptation using discriminative mapping transformsInternational Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2008
- fMPE: Discriminatively Trained Features for Speech Recognition2006
- Improvements in MLLR-Transform-based Speaker Recognition2006
- Improvements to fMPE for discriminative training of features2005
- Discriminative linear transforms for feature normalization and speaker adaptation in HMM estimationIEEE Transactions on Speech and Audio Processing, 2005
- A compact model for speaker-adaptive training2002