Robust Speaker Verification with Stochastic Matching
In today’s telecommunications environment, which includes wireless, landline, VoIP, and computer networks, the mismatch between training and testing environments poses a big challenge to speaker authentication systems. In Chapter 8, we addressed the mismatch problem from a feature extraction point of view. In this chapter, we address the problem from an acoustic modeling point of view. These two approaches can be used independently or jointly.
KeywordsHide Markov Model Equal Error Rate Speaker Verification Test Utterance Feature Extraction Point
Unable to display preview. Download preview PDF.
- 4.Li, Q., Parthasarathy, S., Rosenberg, A. E.: “A fast algorithm for stochastic matching with application to robust speaker verification,” in Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing (Munich), pp. 1543–1547, April 1997Google Scholar
- 7.Parthasarathy, S., Rosenberg, A. E.: “General phrase speaker verification using sub-word background models and likelihood-ratio scoring,” in Proceedings of ICSLP-96 (Philadelphia), October 1996Google Scholar
- 9.Rosenberg, A. E., Lee, C.-H., Soong, F. K.: “Cepstral channel normalization techniques for HMM-based speaker verification,” in Proceedings of Int. Conf. on Spoken Language Processing (Yokohama, Japan), pp. 1835–1838, 1994Google Scholar
- 10.Sankar, A., Lee, C.-H.: “A maximum-likelihood approach to stochastic matching for robust speech recognition” IEEE Transactions on Speech and Audio Processing 4, 190–202 (1996)Google Scholar
- 11.Surendran, A. C.: Maximum-likelihood stochastic matching approach to non-linear equalization for robust speech recognition. PhD thesis, Rutgers University, Busch, NJ, May 1996Google Scholar