Skip to main content

SPEECH DATABASES FOR CREATION OF VOICE DRIVEN TELESERVICES

Exploitable results

The project is producing spoken language resources (speech databases) for use in developing teleservices such as information services, transaction services and other call processing services. Speaker identification/verification facility has been developed which is functionally integrated with the spoken language interface in financial/telecom services. The results of the project will provide protection for services offered over the public telephone network, such as telephone banking, which require a high level of security and protection against fraud. URL: speechdat.phonetik.uni-muenchen.de/
SpeechDat-CAR is producing large-scale spoken language resources (speech databases) for developing in-car voice driven information services and control systems. SpeechDat-CAR extends the SpeechDat project with speech recordings made in noisy car environment and with hands-free operation of GSM. These allow hands-free operation of the increasing complicated array of in-car devices, improving safety and opening up the potential for greater in-car use of network services, such as traffic, weather, routing, and business and personal information. The resources will cover 300 speakers, 9 European languages and be suited to development of a range of applications, such as operation of telephone, navigation aids, entertainment unit, and other command and control of devices. The databases will be offered for purchase by third parties in the year 2002, thereby widening the exploitation to non-consortium members.