A prototype of distant-talking interface for control of interactive TV

This paper aims to describe goals, challenges, and main achievements of DICIT EC project, which addressed the development of a multi-modal user-friendly interface for control of SetTopBox, TV and related services. The interface includes a microphone array to support distant-talking voice input with multiple active speakers. The front-end processing component feeds a chain including speech recognition, natural language understanding, and spoken dialogue management components. The resulting prototype was replicated at several sites and evaluated by 172 users. Results of this campaign showed effectiveness of the adopted solution as well as potential for future development of real smart-space applications.