Jul 14 2015
The DIRHA scientific project coordinated by Trento-based Fondazione Bruno Kessler was judged “excellent” by the EU.
Based on advanced voice recognition and audio signal processing technology, the prototype was mainly designed for people with mobility impairments.
"Excellent progress. The project has fully achieved its objectives and technical goals for the period, and has even exceeded expectations." reads the official report by the European Commission recently received, after the conclusion of the DIRHA (Distant-speech Interaction for Robust Home Applications) scientific project coordinated by Fondazione Bruno Kessler in Trento.
The prototype was especially designed for people with disabilities and allows them to use their voice as a remote control to access services and operate a number of home devices. Adjusting room temperature, opening or closing roller shutters, turning lights on and off, opening or closing doors and windows, selecting and playing music. For these and other applications, DIRHA uses a network of microphones placed throughout the room so it can pick up voice commands, understand them, and then transmit them to a home automation system module, through which the corresponding action is executed.
The project lasted three years, received a 3 and a half million Euro funding by the European Union, 900,000 of wich awarded to FBK, and saw the participation of Athena IAMU RC (Greece) and INESC-ID (Portugal) research centers, the University of Graz (Austria), and companies such as DomoticArea (Rovereto), STMicroelectronics (Milan) and NEW AMUSER (Torino).
The work was coordinated by Maurizio Omologo, Head of the SHINE (Speech-acoustic scene analysis and interpretation) Research Unit at Fondazione Bruno Kessler’s ICT (Information and Communication Technology) Center, focusing on this type of advanced technology.
"DIRHA - says Omologo - represents the natural evolution of the previous DICIT European project, also coordinated by FBK, and through which in 2009 a TV-voice interaction system was developed. In the context of DIRHA, during the past three years, we have been able to achieve important progress, based on which we now have technology that can be used for a voice assistant for household environments, which has some important advantages compared to what is currently available on the market. "
The DIRHA system can be configured either through traditional microphones, or thorugh MEMS (microelectromechanical) microphones in the order of magnitude of a few millimeters, small enough to blend with the furnishings of the home. It can be set to operate 24 hours a day, always listening and waiting for any requests from the user, and there is no need to press any buttons to activate it. It can recognize the occupants’ voices from other background noises and interact with the user only when needed, for example by asking request feedback questions.
One can talk to DIRHA even at several meters from the microphones and it does not need Internet connection, as opposed to most of the solutions available on the market, which require access and audio transmission to a remote service, resulting in limitations in terms of privacy. Moreover, thanks to DIRHA, multiple users located in different rooms of the house can interact with the system simultaneously and independently.
At this time, the prototype is available in Italian, German, Portuguese and Greek. The English version is expected to be added shortly.