Audio and text synchronization for TV news subtitling based on Automatic Speech Recognition

Abstract

The access to the Information Society for deaf and hard of hearing people is a major issue that must be addressed. Subtitling of broadcast news programs is a crucial application that must be incorporated by television companies in order to allow this accessibility and not only for deaf people but for non-impaired ones in noisy and/or crowded places. However, live subtitling is a high cost operation in terms of high qualification human resources and thus, money if high precision is desired. Even assisted systems (that are partially automatic needing human assistance) could result too expensive compared with a fully automatic system. Automatic Speech Recognition researchers can help to perform this task saving both time and money developing systems that delivers subtitles fully synchronized with the announcer speech without human assistance. In this paper we present a real-time system for automatic subtitling of broadcast news programs in Spanish based on the news redaction computer systems (NRCS) texts and an automatic speech recognition engine to allow proper temporal alignment.