A Study of the Capability of Grammatical Analysis to Improve Accuracy in Continuous Speech Recognition for Command and Control.
Interim scientific rept.,
CENTER FOR COMMUNICATIONS RESEARCH INC ROCHESTER N Y
Pagination or Media Count:
Effort has been concentrated on the development of the major components of the complete speech understanding system. The acoustic analysis and segmentation procedures have been run on a modest sample of speech. The segment classification program has been run. The lexical processing has been run extensively with errorful alphabetic strings substituting for phonetic strings. The syntax model is complete and a beginning has been made on the semantic model necessary for processing inputs over limited subject matter. A major effort was expended in integrating the various components which have individually reached maturity. The analysis indicates the sensitivity of the total solution to the segment classification and lexical processing. Modified author abstract
- Command, Control and Communications Systems