Title :
Possible neural networks to detect the glottal wave in speech
Author :
Beddoes, Michael P.
Author_Institution :
British Columbia Univ., Vancouver, BC, Canada
Abstract :
The glottal wave is the forcing wave which passes mainly through the vocal tract to the lips at which the speech wave is emitted. The speech wave and the glottal wave are periodic for “voiced” sounds and not periodic for “unvoiced” sounds. The author suggests that a pair of neural networks can be used to unravel at least one form of the glottal wave directly from the speech wave and to provide estimates of the pitch frequency FO. The paper discusses the following topics: motivation including a brief review and why neural networks can be used; two experimental versions of the glottal wave; details of proposed neural networks; and storage and capacity of the neural network
Keywords :
acoustic signal processing; neural nets; speech processing; forcing wave; glottal wave detection; neural networks; pitch frequency; speech wave; vocal tract; Convolution; Equations; Frequency estimation; Humans; Intelligent networks; Lips; Lungs; Neural networks; Nose; Speech processing;
Conference_Titel :
Systems, Man and Cybernetics, 1995. Intelligent Systems for the 21st Century., IEEE International Conference on
Conference_Location :
Vancouver, BC
Print_ISBN :
0-7803-2559-1
DOI :
10.1109/ICSMC.1995.537846