SITA – Speech Intelligibility Transformation & Autocorrection

Better sound, less noise! That is the slogan of the SITA project funded by the Federal Ministry of Education and Research. The acronym stands for Speech Intelligibility Transformation & Autocorrection. The project is concerned with the improvement of speech intelligibility, especially in the area of multimedia. What does this mean in practice? Many viewers know the problem: Speech in a TV programme can be difficult to understand due to background music or what are known as atmospheric effects. SITA addresses the main causes for poor speech intelligibility along the whole distribution chain of a film and aims to eliminate technological barriers for people with and without hearing impairments. In a nutshell, the project focuses on all applications and hearing situations, commencing from the very start of a film production. A cross-platform software system for sound improvement in the area of media was already developed by the partners in the framework of the “SI4B – Speech Intelligibility for Broadcast” project. This is now being expanded here.

The algorithms developed at Fraunhofer IDMT allow objective measurement as well as automatic improvement of speech intelligibility for specific target groups. The system is intended not only to help broadcasting companies, production firms, distributors and content providers to give end users a better listening experience. Service providers in the media sector and manufacturers in the area of home entertainment and consumer electronics should also profit.