SpeechDat-Car: Towards a collection of speech databases for automotive environments
Henk van den Heuvel, Antonio Bonafonte, Jérôme Boudy, Simon Dufour, P Lockwood, Antonio Moreno‐Muñoz, Gaël Richard · 1999
The SpeechDat-Car project is a 4 th framework EC project in the Language Engineering programme. It aims at collecting a set of nine speech databases to support training and testing of robust multilingual speech recognition for in-car applications. The consortium participants are car manufacturers, telephone communications providers, and universities. This paper describes the background of the project, its organisation, and the design of the databases in terms of contents, speaker and environment coverage. It further addressses the recording platforms, the validation scenario, and the links with other projects. 1. INTRODUCTION Speech recognition technology has reached a point where it is sufficiently robust to operate in relatively noisy environments, if sufficient efforts are made to cope with the noise problem. A notoriously noisy environment is the car cockpit. Here, speech recognition presents an attractive option. For instance, operating devices like telephones in th...