Method and apparatus for evaluating the accuracy of a speech recognition system

作者: Kerry A. Ortega

DOI:

关键词:

摘要: A method and system for evaluating the accuracy of a computer speech recognition counts indexes total number words dictated corrected. The corrections are tallied after being made in correction window include contained an alternative list as well input by user within stored word database. processor calculates approximate ratio correct to dictated. An is calculated each dictation session overall all sessions combined. also keeps individual times corrected were alternate lists or not database uses these calculate additional values.

参考文章(12)
James R. Lewis, Maria Smith, Automatically updating language models Journal of the Acoustical Society of America. ,vol. 109, pp. 2553- ,(1998)
Joseph H Miglietta, Douglas Holt, Michael K Davis, Word tagging and editing system for speech recognition ,(1995)
Xuedong D. Huang, Hsiao-Wuen Hon, Li Jiang, Method and system for correcting misrecognized spoken words or phrases ,(1996)
Erik Bunce, Harley Rosnow, Douglas Gerard, Speech recognition and editing system ,(1997)
Barbara E. Ballard, Kerry A. Ortega, Marking and deferring correction of misrecognition errors ,(1998)
Dean Sturtevant, Daniell Stevens, Joel M. Gould, Charles E. Ingold, Michael J. Newman, Allan Gold, David Abrahams, Robert Roth, Error correction in speech recognition ASAJ. ,vol. 109, pp. 30- ,(2001)