11-2 Experiment of Real-Time Captioning for Broadcast News Using Speech Recognition of Direct Program Sound and Re-Spoken Utterances

Abstract

This paper describes an experiment of a real-time captioning system for broadcast news by using hybrid speech recognition of direct program sound and rephrased utterances by a re-speaker. The result shows that the system could be used for captioning typical news programs operated by one or two error correctors and a re-speaker.