The new EU project “LISTEN” is working on creating a robust, hands-free speech control interface for home systems. The partners from Germany, Italy and Greece will develop a smart-home-specific natural voice interface to web services. Moreover, they want to bridge the gap between microphone systems, signal processing and speech recognition.
Today, it is becoming more and more attractive and affordable for everyone to enhance the home environment with software to remotely control heating or cooling, lighting, media and communication – these smart home functionalities are an essential aid for the elderly and people with disabilities. They provide assistive control of important everyday activities and, in case of emergencies, can even save lives. However, smart home functionalities still cannot be accessed in a satisfying manner via a natural, easy to use, interface: the voice.
Therefore speech technology and signal processing experts are working on the new EU project “LISTEN” - “Hands-free Voice-enabled Interface to Web Applications for Smart Home Environments”. They will design a system enabling robust hands-free large-vocabulary voice-based access to Internet applications in smart homes. Two crucial components of this system are large recognition vocabularies and suitable microphone arrays. This will allow you to control smart-home web-enabled functionalities, and also access specific Internet applications, using your voice. “LISTEN” is part of the Marie Skłodowska-Curie Research and Innovation Staff Exchange (RISE) program, in the HORIZON 2020 framework. The four-year project started on June 1, 2015. The partners are: the Foundation for Research and Technology - Hellas (FORTH), Greece (project coordinator), the Heidelberg-based EML European Media Laboratory GmbH, Germany, RWTH Aachen, Germany, and Cedat 85, Italy.
One goal of the project is to develop a system with which the user can interact by only speaking. Therefore, a truly hands-free operation of the voice interface is obligatory. Users will not have to turn towards a microphone or other device, or wear a headset. This heavy task requires a robust hands-free speech capturing system operating as a wireless acoustic sensor network, which will be specifically designed for the smart home. The project will include both software and hardware components that have to perform in real-time.
Another goal is that the user may access web applications and control smart home functions within the same user interface. For web applications like web search, email dictation and access to social networks, a large-vocabulary automatic speech recognition system will be required. “We will optimize our speech recognition system for accessing these web services and, at the same time, to control smart home automation functionalities”, says Dr. Siegfried Kunzmann, R & D manager at the EML European Media Laboratory (EML). “We will also support multiple languages.” The project´s focus is on English, Greek, Italian, and German, to demonstrate the system’s easy adaptation to further languages.
With these two goals, LISTEN pushes the boundaries of the current state-of-the-art. The partners are seeking to bridge the gap between the acoustic sensors and the automatic speech recognition research communities. They bring in a lot of expertise from both sides, thus strengthening interdisciplinary research and the exchange of knowledge between universities and companies on a European level.
The Institute of Computer Science, of the Foundation for Research and Technology - Hellas (FORTH), Heraklion, Greece, serves as project coordinator. They bring to the project their expertise in acoustic sensor networks and microphone array signal procesing. The Heidelberg-based EML European Media Laboratory GmbH (Germany) will provide their speech recognition system and refine their tools for the project. The RWTH Aachen (Germany) speech recognition group will bring in its long-time experience and newest technologies in core speech recognition research. Cedat 85 (Italy) brings to LISTEN its industrial expertise on building specific applications and systems that incorporate speech recognition technologies.
For further information, please contact:
Athanasios Mouchtaris, Associate Professor of Computer Science
Institute of Computer Science, Foundation for Research and Technology - Hellas (FORTH)
N. Plastira 100, GR-700 13 Heraklion, Crete
LISTEN website: http://www.listen-project.eu
Foundation for Research and Technology - Hellas (Greece)is one of the largest research centers in Greece with well organized facilities, highly-skilled personnel and a reputation as a top-level research foundation worldwide. In LISTEN, FORTH participates through its Institute of Computer Science (ICS), and more specifically the Signal Processing Laboratory (SPL) which has significant expertise on acoustic sensor networks, sound source localisation and beam forming, audio signal modelling and coding, distributed processing, and communications.
The EML European Media Laboratory GmbH (Germany) is a private IT enterprise established by Klaus Tschira, one of the founders of the SAP AG software company. In accordance with its motto “Experience IT – Intuitive Technology” EML is successfully pursuing research and development in the fields of human-computer interaction and automatic speech processing.
The RWTH Aachen (Germany) speech recognition group is one of the most prestigious in this area worldwide, with decades of experience in core speech recognition research, including statistical model development, natural language processing, and applications development.
Cedat85 (Italy) has a long history of industrial expertise on building specific applications of speech recognition systems, including system integration and evaluation of speech communication systems.