MENU

A platform for the applications of speech technologies on smartphones for the languages of the danube region

The project is aimed at the development of a software platform for a range of voice applications for smartphones in the languages of the danube region. A range of user categories, including national minorities and the disabled, will be able to use efficient voice interfaces in their mother tongues.

The field of speech technology applications has been significantly extended by the emergence of smartphones, which require an efficient voice user interface, e.g. for RSS feed reading, SMS/tweet/email reading or name dialing. The utility of such applications and ultimately their success depends on their ability to support individual languages, and this is one of the problems that will be treated within this project. The project will give rise to a synergy and the exchange of experience, resources and knowledge between partners in order to develop a platform for a range of useful applications of speech technologies on mobile devices using different operating systems, adaptable to the needs of different types of users, from different languages to the persons with complex communication needs. Joint activities on the improvement of the expressiveness and flexibility of text-to-speech synthesis (TTS) will lead to its easier application in a wider range of services. Similarly, increased accuracy and robustness of automatic speech recognition (ASR) in particular domains of interaction, as well as its adaptation to the voice of its user, will increase the range of its application. Many people whose communication capabilities are severely reduced for any of a variety of reasons, use augmentative alternative communication (AAC) instead of conventional means of communication between humans, i.e. speech or written language. They prefer a speech enabled application converting strings of visual symbols into meaningful sentences. The platform based on ASR, TTS and AAC will be developed within a coordinated effort of all project partners, each participating country will adapt it to its language(s), and the project will thus have a significant impact on all multilingual regions within the Danube Region. Apart from increased competitiveness of participating SMEs, the project will bring a range of useful speech applications to end users in multinational societies of the region. The government and news web sites offering information in multiple languages, including minority languages, will provide support to the disabled, but also to many other people who prefer listening to news instead of reading them on small screens of their mobile devices. Commercial web sites financed by commercial ads will be offered a new type of advertisement, through short spoken messages accompanying articles, such as "Reading of this message was provided by [name of sponsor]". Among others, the platform will provide state and local governments with the possibility to broadcast important messages and alerts to mobile phones of persons at particular locations in their mother tongue, including minorities, tourists, as well as in different modalities for the disabled. The proposed partners in the consortium include the following: - The Faculty of Technical Sciences, University of Novi Sad - FTS-UNS (SR), the regional leader in the research and development of speech technologies for Serbian and Croatian, which has developed high-quality TTS and ASR systems for these languages. The EUREKA project will also benefit from the results of ongoing technological projects of FTS-UNS, related to the development of spoken dialogue systems in South Slavic languages. - "AlfaNum Ltd." (SR) and "HSM-informatika Ltd." (HR) have cooperated with FTS-UNS on the development of speech technologies and their applications, used all over SR, HR, ME, and BA. AlfaNum has recently developed the application "Voice Assistant" for voice interaction with smartphones based on ASR, and is working on speaker adaptation and the development of expressive TTS systems. The research group from HSM has focused on ICT systems for symbol based AAC and accessible ICT services. It implemented the EU-funded project “ICT Competence Network for Innovative Services for Persons with Complex Communication Needs”. - The Department of Speech Analysis and Synthesis of the Institute of Informatics in Bratislava (SK) is experienced with speech processing in Slovak and partly also in Hungarian and Serviko Romani (spoken by the Slovak Roma minority), with a rich experience in expressive TTS (voice alerts in emergency situations, dramatic piece reader, etc.). - SpeechTex Ltd. (HU) is a startup company focused on speech technology. Although it was founded in 2014, its founders are experienced in speech technology, especially ASR. The ASR engine and know-how developed at SpeechTex are used in call center speech analytics projects, e.g. in the largest Hungarian bank, and in automated closed captioning solutions for Hungarian national TV. - The Speech Communication and Smart Interactions Lab of BME-TMIT (HU) has over 30 years of multilingual speech technology expertise. Besides basic research it has developed professional, real-time telecommunication services. Their ProfiVox-Corpus TTS technology has been installed at more than 60 Hungarian railway stations for audio passenger information.
Acronym: 
DANSPLAT
Project ID: 
9 944
Start date: 
01-01-2016
Project Duration: 
36months
Project costs: 
770 000.00€
Technological Area: 
ELECTRONICS, IT AND TELECOMS TECHNOLOGY
Market Area: 
SERVICES

Raising the productivity and competitiveness of European businesses through technology. Boosting national economies on the international market, and strengthening the basis for sustainable prosperity and employment.