Presentation	1998/11/20 Speech recognition using Recurrent Neural Prediction Model Toru Uchiyama, Haruhisa Takahashi,
PDF Download Page	PDF download Page Link
Abstract(in Japanese)	(See Japanese page)
Abstract(in English)	We propose a new speech recognition model, called"Recurrent Neural Prediction Model, RNPM", via Recurrent Neural Network(RNN). RNN is advantageous when acquiring the capability of categorizing temporal sequences by learning. Thus, it can be used as a temporal sequence predictor, as an application to the speech recognizer. We apply RNN to "Neural Prediction Model, NPM"(Iso 1989)in the hope that it can improve the learning ability and the generalization. Especially, a new RNN architecture, which is based on Jordan and Elman's network, is used for RNPM. We performed speaker independent isolated digit recognition through computer simulation, which attained perfect recognition for unknown examples.
Keyword(in Japanese)	(See Japanese page)
Keyword(in English)	Recurrent Neural Network / Prediction Model / Digit Recognition
Paper #	SP98-94
Date of Issue

Conference Information
Committee	SP
Conference Date	1998/11/20(1days)
Place (in Japanese)	(See Japanese page)
Place (in English)
Topics (in Japanese)	(See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To	Speech (SP)
Language	JPN
Title (in Japanese)	(See Japanese page)
Sub Title (in Japanese)	(See Japanese page)
Title (in English)	Speech recognition using Recurrent Neural Prediction Model
Sub Title (in English)
Keyword(1)	Recurrent Neural Network
Keyword(2)	Prediction Model
Keyword(3)	Digit Recognition
1st Author's Name	Toru Uchiyama
1st Author's Affiliation	Department of Communications and Systems, The University of Electro-Communications()
2nd Author's Name	Haruhisa Takahashi
2nd Author's Affiliation	Department of Communications and Systems, The University of Electro-Communications
Date	1998/11/20
Paper #	SP98-94
Volume (vol)	vol.98
Number (no)	424
Page	pp.pp.-
#Pages	7
Date of Issue