Yet another acoustic representation of speech sounds
Author:
- 28 September 2004
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 1, 585
Abstract
This paper proposes yet another representation of speech sounds. The proposed speech modeling can remove both multiplicative and linear transformational distortion from speech theoretically. It means that speech sounds are represented without being affected by any static distortion inevitably involved in production, encod- ing, transmission, decoding, and hearing processes, such as differ- ences in vocal tract length, gender, age, microphone, room, line, auditory characteristics, and so on. The method acoustically mod- elsnot individualphones but theirentire system, where only acous- tic interrelation embedded in all the kinds of phones is focused. Since the method provides us with no absolute acoustic properties of phones, it cannot recognize or synthesize even a single phone. On the contrary, the proposed method is shown to be able to be ap- plied to pronunciation assessment effectively and reliably, where the proficiency of pronunciation is estimated without using acous- tic models of the individual phones directly in the matching.Keywords
This publication has 2 references indexed in Scilit:
- Vocal tract normalization equals linear transformation in cepstral spaceIEEE Transactions on Speech and Audio Processing, 2005
- Maximum likelihood linear regression for speaker adaptation of continuous density hidden Markov modelsComputer Speech & Language, 1995