Paper: Real-Time Correction of Closed-Captions

ACL ID P07-2029
Title Real-Time Correction of Closed-Captions
Venue Annual Meeting of the Association of Computational Linguistics
Session System Demonstration
Year 2007

Live closed-captions for deaf and hard of hearing audiences are currently produced by stenographers, or by voice writers us- ing speech recognition. Both techniques can produce captions with errors. We are cur- rently developing a correction module that allows a user to intercept the real-time cap- tion stream and correct it before it is broad- cast. We report results of preliminary ex- periments on correction rate and actual user performance using a prototype correction module connected to the output of a speech recognition captioning system.