Title :
An associative memory approach to isolated-utterance speech recognition
Author :
Ervin, Thomas C., II ; Kim, Jung H.
Author_Institution :
Dept. of Electr. Eng., North Carolina A&T State Univ., Greensboro, NC, USA
Abstract :
A method for achieving isolated-utterance speech recognition using an associative recall algorithm is presented. The feature analysis methods used, as well as the implementation of the associative memory scheme, are highlighted. Preliminary experimental results have proved this method to be quite robust in the recognition aspect. Most of the experiments were performed with speaker-dependent tests and proved to be quite successful with a very limited vocabulary
Keywords :
Hopfield neural nets; content-addressable storage; convergence; correlation methods; learning by example; linear predictive coding; self-organising feature maps; speech recognition; associative memory scheme; associative recall algorithm; feature analysis; isolated-utterance speech recognition; speaker-dependent tests; Associative memory; Autocorrelation; Delay estimation; Linear predictive coding; Neural networks; Signal analysis; Spectral analysis; Speech analysis; Speech processing; Speech recognition;
Conference_Titel :
System Theory, 1993. Proceedings SSST '93., Twenty-Fifth Southeastern Symposium on
Conference_Location :
Tuscaloosa, AL
Print_ISBN :
0-8186-3560-6
DOI :
10.1109/SSST.1993.522838
