PDF Basket
It is becoming increasingly affordable to automate homes, enabling the remote control of heating and air-conditioning, lighting and even online communication. However, the elderly and those who are physically challenged may not always be able to benefit from this technology if it is not easy to use.
The EU-funded LISTEN project tackles this challenge by enabling people to operate a wide range of functions in their smart homes, including web-linked appliances such as lamps or blinds, by simply using their voices. In a bid to better meet the needs of seniors and the disabled, who may often be homebound and isolated, the system also makes it possible to dictate emails, access social media and search the web – all without typing a single word.
‘Far-field voice control is not just nice to have but essential for people with disabilities and the elderly who either have difficulty or may not be able to access web-based services that keep us connected to information and to friends and family,’ says LISTEN project coordinator Athanasios Mouchtaris of the Institute of Computer Science at the Foundation for Research and Technology Hellas in Greece.
Life-saving technology
LISTEN has seen key results both in wireless acoustic sensor networks (WASNs) for voice enhancement regardless of the speaker’s location and orientation in a smart home, as well as in large-vocabulary speech recognition for web-based services. In this respect, research was done in both directions jointly rather than independently of one another; the results have been published in peer-reviewed international journals.
On a practical level, users do not have to wear a headset or even turn towards a microphone in order to communicate with the outside world, which is important for people who are less mobile or visually impaired. This ease of use could also be potentially life-saving in the event of an emergency when being able to quickly call for help is key.
Most of the companies developing smart-home voice interface systems are US-based, so LISTEN offers EU-based academics and industry an opportunity to work on cutting-edge technology for voice interaction and to gain market momentum. While the plan is to eventually commercialise the results of LISTEN research through industrial partners actively involved in the expanding area of voice interface, US patent applications have been filed.
Addressing privacy concerns
The vast majority of voice-based interfaces developed thus far have been Cloud-based, resulting in consumer concerns about potential privacy breaches. LISTEN, which includes both software and hardware components, takes a different tack.
‘A potential solution to such privacy concerns could be a system that operates on a local level and does not send any data outside the user’s environment, which is a key reason for LISTEN to concentrate on this approach,’ says Mouchtaris.
The project has placed emphasis on the training of young researchers, funding their involvement in research activities and participation in major international conferences and workshops. Its technology results have been presented to high-school students to teach them about the challenges and rewards associated with being a computer scientist. LISTEN also co-funded two international summer schools focused on speech technologies.
LISTEN received funding through the EU’s Marie Skłodowska-Curie actions programme.