Speech Understanding Systems

Abstract
The elements of the mid-term system were integrated successfully and it was demonstrated in April 1974. System performance in a controlled test of 116 utterances was 74 percent correct for a 248-word vocabulary, and 50 percent correct for a 500-word vocabulary. Software to support packet speech communication via CVSD modulation techniques over the ARPA computer network has been completed and checked out as far as possible. This is the last SATS concerning speech understanding systems. Future SATS will be concerned with work on packet speech communication.

This publication has 0 references indexed in Scilit: