SPEECH RECOGNITION SYSTEM AND METHOD
 Inventors: Ben Warren, Saratoga; George M.
White, Los Altos; Rick Parfitt, Los
Gatos; Peter Deng, San Jose, all of
 Assignee: Threshold Technology, Inc., Delran, N.J.
 Appl.No.: 240,273
 Filed: Mar. 4,1981
 Int. C1.3 G10L 1/00
 U.S. CI 179/1 SD; 364/513
 Field of Search 179/1 SD, 1 SM, 1 SC;
340/146.3 WD, 146.3 AQ; 364/513
 References Cited
U.S. PATENT DOCUMENTS
4,069,393 1/1978 Martin et al 179/1 SD
4,100,370 7/1978 Suzuki et al 179/1 SD
FOREIGN PATENT DOCUMENTS
1225242 3/1971 United Kingdom .
J. Shearme et al., "Some Experiments with a Word
Recognition System," IEEE Trans, on Audio and Elec, Jun. 1968, pp. 256-261.
S. Nakagawa et al., "A Realtime Spoken Word Recognition System", Computer Controls, May-Jun., 1978, pp. 63-71.
Clapper, "Connected Word Recognition System", IBM
Tech. Bull., Dec. 1969, pp. 1123-1126.
Primary Examiner—Emanuel S. Kemeny
Attorney, Agent, or Firm—Martin Novack
An apparatus for receiving spoken input training words and for subsequently recognizing a spoken input command word. For each training word, and for the command word, there is stored a time dependent array and a cumulative feature non-time-dependent vector. The command word array is compared, member-by-member, with each training word array. Also, the command word vector is compared, feature-by-feature, with each training word vector. The results of both comparisons are used for selecting the training word which most closely matches the command word.
24 Claims, 10 Drawing Figures