Title :
A comparative study of spectral mapping for speaker adaptation
Author :
Nakamura, Satoshi ; Shikano, Kiyohiro
Author_Institution :
Sharp Corp., Nara, Japan
Abstract :
A comparative study of speaker adaptation using neural network spectral mapping and fuzzy vector quantization (VQ)-based spectral mapping is described. The speaker adaptation experiments were carried out using a database of 216 phonetically balanced words uttered by three speakers. The accuracy of spectral mapping is measured and evaluated by interspeaker spectral distortion. The results show the fuzzy VQ-based spectral mapping algorithm to be about 6% better in spectral distortion than nonlinear spectral mapping by a feedforward neural network. An investigation of the actual spectrogram shows that the fuzzy VQ-based spectral mapping works better than the neural network mapping
Keywords :
adaptive systems; neural nets; spectral analysis; speech analysis and processing; speech recognition; feedforward neural network; fuzzy vector quantization; interspeaker spectral distortion; nonlinear mapping; phonetically balanced words; speaker adaptation; spectral mapping; speech recognition; Approximation algorithms; Databases; Distortion measurement; Feedforward neural networks; Fuzzy neural networks; Laboratories; Neural networks; Nonlinear distortion; Piecewise linear techniques; Spectrogram; Speech recognition; Speech synthesis; Telephony; Vector quantization;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1990. ICASSP-90., 1990 International Conference on
Conference_Location :
Albuquerque, NM
DOI :
10.1109/ICASSP.1990.115562
