Speech pattern recognition using pattern recognizers and classifiers

United States Patent im

Linggard et al.

US005794190A til] Patent Number: [45] Date of Patent:

5,794,190 Aug. 11, 1998

[54] SPEECH PATTERN RECOGNITION USING PATTERN RECOGNIZERS AND CLASSIFIERS

[75] Inventors: Robert Linggard. Suffolk; Philip

Charles Woodland, Cambridge; John
Edwrd Talintyre. Essex; Samuel
Gavin Smyth. Suffolk, all of England

[73] Assignee: British Telecommunications public
limited Company. London. England

[21] Appl. No.: 338,158

[22] Filed: Nov. 9, 1994

Related U.S. Application Data

[63] Continuation of Ser. No. 955,730, filed as PCT/GB91/00668 Apr. 26, 1991 published as W091/16687 Oct. 31, 1991, abandoned.

[30] Foreign Application Priority Data

Apr. 26, 1990 [GB] United Kingdom 9009425

[51] Int CI.6 G10L 9/00

[52] U.S. CI 704/232; 704/239

[58] Field of Search 395/2. 2.11. 2.41.

395/2.68, 21. 23, 24; 381/42. 43; 704/200, 202. 232. 259. 240. 243. 244. 239. 245

[56] References Cited

U.S. PATENT DOCUMENTS

5,031^17 7/1991 Nishimura 381/43

5,083^85 1/1992 Shimaetal 395/24

5^28,087 7/1993 Bickerton 381/43

5,317,673 5/1994 Cohen etal 395/2.41

FOREIGN PATENT DOCUMENTS

WO 89/02134 3/1989 WIPO .

OTHER PUBLICATIONS

IEEE First International Conference On Neural Networks, 21-24 Jun. 1987, San Diego. CA.. R.P. Lippmann et al.: "Neural-Net Classifiers Useful for Speech Recognition", pp. IV 417-425.

IEEE International Conference On Neural Networks, 24-27
Jul. 1988. San Diego. CA... IEEE. (New York). C.-H. Sung
et al: 'Temporal Pattern Recognition", pp. 1-689—1-696.
British Telecom Technology Journal, vol. 6. No. 2. Apr.
1988. N. McCulloch et al.: "Multi-Layer Perceptions
Applied to Speech Technology", pp. 131-139.
IEEE Transactions On Acoustics. Speech. And Signal Pro-
cessing, vol. ASSP-27, No.4. Aug. 1979. IEEE (New York).
L.R. Rabiner et al.: "Speaker-Independent Recognition of
Isolated Words Using Clustering Techniques", pp. 336-349.
British Telecom Technology Journal, vol. 6, No. 2. Apr.
1988, S.J. Cox: "Hidden Markov Models for Automatic
Speech Recognition: Theory and Application", pp. 105-115.
Proceedings Of The JEEE, vol. 77. No. 2. Feb. 1989. L.R.
Rabiner: "a Tutorial on Hidden Markov Models and
Selected Applications in Speech Recognition," pp. 257-286.
R. Lippmann. "A Critical Overview of Neural Network
Pattern Classifiers." Neural Networks for Signal Processing,
Sep. 30-Oct. 1. 1991. pp. 266-275.
W. Huang, R. Lippmann, and B. Gold, "A Neural Net
Approach to Speech Recognition," ICASSP '88, Apr. 4-11,
1988. pp. 99-102.

(List continued on next page.)

Primary Examiner—Allen R. MacDonald

Assistant Examiner—Richemond Dorvil

Attorney, Agent, or Firm—Nixon & Vanderhye PC.

[57] ABSTRACT

An apparatus for recognizing predetermined patterns comprises an input for signal data; a pattern matching stage including a plurality of stores, each storing a representation of a pattern to be recognized; a siinilarity measurer for producing, for each representation, a measure of the similarity between the signal data and that representation; and a classification stage for deciding from the similarity measures which, if any, of the patterns to which the representations correspond is present in the signal data, and for indicating the presence of any such pattern. The classification stage makes its determination based on all the similarity measures after having been (rained using information relating to known patterns and the similarity measurer from the pattern matching stage produced by inputting known patterns.

23 Claims, 2 Drawing Sheets

OTHER PUBLICATIONS

S. Austin. G. Zavaliagkos, J. Makhoul. and R. Schwartz, "A Hybrid Continuous Speech Recognition System Using Semental Neural Nets with Hidden Markov Models." Neural Networks for Signal Processing. Sep. 30-Oct. 1. 1991, pp. 347-356.

S. Austin, G. Zavaliagkos. J. Makhoul, and R. Schwartz.

"Speech Recognition Using Segmental Neural Nets."

ICASSP '92. Mar. 23-26, 1992. pp. 625-628.

S. Austin. G. Zavaliagkos. J. Makhoul. and R. Schwartz,

"Continuous Speech Recognition Using Segmental Neural

Nets," Jun. 7-11. 1992, pp. 314-319.

R. Lippmann and E. Singer, "Hybrid Neural-Network/

HMM Approaches to Wordspotting." ICASSP '93. Apr.

27-30. 1993. pp. 565-568.

E. Singer and R. Lippmann. "A Speech Recognizer Using Radial Basis Function Neural Networks in an HMM Framework." ICASSP '92, Mar. 23-26. 1992, pp. 629-632.

Gao et al, "HMM Based Warping in Neural Networks." ICASSP '90. Apr. 3-6. 1990. pp. 501-504.

Prinslou et al, "A Hidden Markov Model For Acoustic Syllabification and Processing Class Labelling In Continous Speech." Comsig 89, Jun. 23. 1989.

Ma et al.. 'TDNN Labeling For A HMM Recognizer," ICASSP '90. Apr. 3-6, 1990. pp. 421^123.

Howard et al.. "Two Level Recognition of Isolated Word Using Neural Word." Artificial Neural Nets, 1989. Oct. 16-18, 1989, pp. 90-94.