Analog LSI design of data clustering/labeling method in speech recognition system

This report first introduces a new speech recognition system. The system is designed with software and hardware mixed systems. In the hardware part, this report shows an analog LSI architecture. This architecture realizes a data clustering/labeling which is used for speech characteristic labeling. The clustering/labeling system which supports a quite simple Japanese vowel labeling mechanism is designed by a 1.2 /spl mu/m CMOS design rule. Its chip area becomes 53700 /spl mu/m/sup 2/.

[1]  Y.K. Muthusamy,et al.  Reviewing automatic language identification , 1994, IEEE Signal Processing Magazine.

[2]  Hervé Bourlard,et al.  Continuous speech recognition , 1995, IEEE Signal Process. Mag..

[3]  Koji Tochinai,et al.  Multi-clustering Network for Data Classification System , 1997 .

[4]  Yoshikazu Miyanaga.  Parallel VLSI architecture for multi-layer self-organizing cellular network , 1993 .

[5]  Yoshikazu Miyanaga,et al.  A robust speech analysis in speech recognition , 2000, WCC 2000 - ICSP 2000. 2000 5th International Conference on Signal Processing Proceedings. 16th World Computer Congress 2000.

[6]  S Mangrulkar,et al.  Artificial neural systems. , 1990, ISA transactions.

[7]  Koji Tochinai,et al.  A self-organized network with a supervised training , 1995, Proceedings of ISCAS'95 - International Symposium on Circuits and Systems.

[8]  Y. Miyanaga,et al.  ARMA digital lattice filter based on a linear prediction theory , 1984 .