Rapid connectionist speaker adaptation

Author

Witbrock, Michael ; Haffner, Patrick

Author_Institution

Sch. of Comput. Sci., Carnegie Mellon Univ., Pittsburgh, PA, USA

Volume

1

fYear

1992

fDate

23-26 Mar 1992

Firstpage

453

Abstract

SVCnet, a system for modeling speaker variability, is presented. Encoder neural networks specialized for each speech sound produce low-dimensionality models of acoustical variation, and these models are further combined into an overall model of voice variability. A training procedure is described which minimizes the dependence of this model on which sounds have been uttered. Using the trained model (SVCnet) and a brief, unconstrained sample of a new speaker´s voice, the system produces a speaker voice code that can be used to adapt a recognition system to the new speaker without retraining. A system which combines SVCnet with a MS-TDNN recognizer is described

Keywords

neural nets; speech recognition; MS-TDNN recognizer; MS-time delay neural network; SVCnet; acoustical variation; connectionist speaker adaptation; low-dimensionality models; model dependence; recognition system; speaker variability; speaker voice code; specialised encoder neural networks; speech sound; training procedure; unconstrained sample; voice variability; Computer architecture; Computer science; Humans; Loudspeakers; Neural networks; Real time systems; Speech coding; Speech recognition; Static VAr compensators; Telephony;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on

Conference_Location

San Francisco, CA

ISSN

1520-6149

Print_ISBN

0-7803-0532-9

Type

conf

DOI

10.1109/ICASSP.1992.225874

Filename

225874