Projects of the Speech analysis and synthesis department VEGA – Scientific Grant Agency of SR and SAS
GAV – Grant Agency for Science
COPERNICUS – European Communities
ASFEU – Agency of SR for EU Structural Funds
SRDA – Agency for research&development support
MAD – Inter-academic agreement

Actual and recent projects:

Security of Air Transport Infrastructure of Europe (SATIE)
Bezpečnosť infraštruktúry leteckej dopravy v Európe
Program: Horizon 2020
Project leader: Ing. Rusko Milan PhD.
ID: 832969
Duration: 1.5.2019 – 31.7.2021
Annotation
Automatic assessment of acute stress from speech.
Automatické hodnotenie akútneho stresu z reči
Program: VEGA
Project leader: Ing. Rusko Milan PhD.
ID:
Duration: 1.1.2018 – 31.12.2020
Annotation

Finished projects:

Relationship between trust and entrainment in speech
Vzťah medzi dôveryhodnosťou a rečovým prispôsobovaním sa
Program: Other
Project leader: Ing. Rusko Milan PhD.
ID: Air Force Office of Scientific Research, FA9550-15-1-0055
Duration: 25.6.2015 – 14.12.2017
Grapheme and Phoneme Frequencies in Slovak
Frekvencie grafém a foném v slovenčine
Program: Bilateral – other
Project leader: Ing. Rusko Milan PhD.
ID: 58s3
Duration: 1.6.2007 – 1.11.2009
Annotation
Global ATM security management (GAMMA)
Globálny bezpečnostný manažment riadenia letovej prevádzky
Program: FP7
Project leader: Ing. Rusko Milan PhD.
ID: FP7-SEC-2012-1
Duration: 1.9.2013 – 30.11.2017
Annotation
Intelligent Language Tutoring System with Multimodal Feedback Functions
Inteligentný systém pre výuku jazykov s multimodálnymi funkciami spätnej väzby
Program: Multilateral – other
Project leader: Ing. Rusko Milan PhD.
ID: 135379-LLP-1-2007-21-DE-KA2-KA2MP
Duration: 1.11.2007 – 31.12.2009
Web page: http://rezo-computer.dyndns.org/euronounce/
Annotation
Nonverbal speech gestures in Slovak in multimodal and multicultural context
Neverbálne rečové gestá v slovenčine v multimodálnom a multikulturálnom kontexte
Program: COST
Project leader: Ing. Rusko Milan PhD.
ID: COST 271/06
Duration: 9.10.2008 – 30.11.2010
Web page: http://www.cost2102.eu/joomla/
Annotation
Automatic meeting speech recognition with application to courtroom recordings transcription.
Automatické rozpoznávanie reči pri komunikácii viacerých osôb s aplikáciou na prepis súdnych pojednávaní.
Program: VEGA
Project leader: Ing. Rusko Milan PhD.
ID: 2/0202/11
Duration: 1.1.2011 – 31.12.2013
Annotation
Automatic subtitling of audiovisual content for people with hearing impairments
Automatické titulkovanie audiovizuálneho obsahu pre osoby so sluchovým postihnutím
Program: SRDA
Project leader: Ing. Rusko Milan PhD.
ID: APVV – 15 – 0517
Duration: 1.7.2016 – 1.7.2018
Automatic transcription of dictate
Automatický prepis diktátu
Program: Other projects
Project leader: Ing. Rusko Milan PhD.
ID:
Duration: 21.4.2009 – 20.4.2011
Annotation
Automated, Voice Controlled Telecommunication Systems and their Applications
Automatizované, hlasom ovládané telekomunikačné systémy a ich aplikácie
Program: Other projects
Project leader: Ing. Rusko Milan PhD.
ID: AV 4 0006 07
Duration: 1.1.2007 – 31.12.2009
Annotation
Electronic services of General prosecution of the SR
Elektronické služby Generálnej prokuratúry SR
Program:
Project leader: Ing. Rusko Milan PhD.
ID: 21110120022
Duration: 1.6.2013 – 1.6.2015
Intelligent Language Tutoring System with Multimodal Feedback Functions
Inteligentný systém pre výuku jazykov s multimodálnymi funkciami spätnej väzby
Program:
Project leader: Ing. Rusko Milan PhD.
ID:
Duration: 1.11.2007 – 31.10.2009
Annotation
Modern speech processing technologies in Slovak utilising speech corpora
Moderné technológie spracovania reči v slovenčine využívajúce rečové databázy II.
Program: VEGA
Project leader: Ing. Rusko Milan PhD.
ID:
Duration: 1.1.2005 – 1.12.2007
Nonverbal speech gestures in Slovak in multimodal and multicultural context
Neverbálne rečové gestá v slovenčine v multimodálnom a multikulturálnom kontexte
Program:
Project leader: Ing. Rusko Milan PhD.
ID:
Duration: 9.10.2008 – 31.12.2009
Web page: http://ui.sav.sk/gestures/
Annotation
Cross-linguistic study of prosodic cues in English, Spanish, and Slovak
Porovnávacia štúdia prozodických signálov v anglickom, španielskom a slovenskom jazyku
Program: Other projects
Project leader: Ing. Rusko Milan PhD.
ID:
Duration: 1.1.2014 – 31.12.2015
Annotation
Speech Technologies for Modern Telecommunication and Information Systems and Services
Rečové technológie pre moderné telekomunikačné a informačné systémy a služby
Program: SRDA
Project leader: Ing. Rusko Milan PhD.
ID: APVV-0369-07
Duration: 1.9.2008 – 31.12.2010
Web page: http://speetis.fei.tuke.sk/
Annotation
Robust Speech Technologies for Information Systems in Slovak and their Diagnostics
Robustné rečové technológie pre informačné systémy v slovenčine a ich diagnostika
Program: VEGA
Project leader: Ing. Cerňak Miloš PhD.
ID: 2/0138/08
Duration: 1.1.2008 – 31.12.2010
Annotation

Štúdia realizovateľnosti projektu „Automatický prepis pojednávaní“
Program: Other projects
Project leader: Ing. Rusko Milan PhD.
ID:
Duration: 1.1.2008 – 31.12.2008
Speaker verification as an add-on element of the Air Traffic Management security
Verifikácia hovoriaceho ako prvok bezpečnosti riadenia letovej prevádzky
Program: VEGA
Project leader: Ing. Rusko Milan PhD.
ID: 2/0197/15
Duration: 1.1.2015 – 31.12.2017
Annotation
(CRISIS)
Výskum a vývoj nových informačných technológií na predvídanie a riešenie krízových situácií a bezpečnosť obyvateľstva
Program: EU Structural Funds Research Development
Project leader: doc. Ing. Hluchý Ladislav CSc.
ID: ITMS 26240220060
Duration: 3.1.2011 – 31.12.2013

Actual and recent projects (annotations):

Security of Air Transport Infrastructure of Europe (SATIE)
Bezpečnosť infraštruktúry leteckej dopravy v Európe
Annotation: The twenty-first century experiments a digital revolution that simplifies flight and cross-border. Digitalizationcontributes to leverage information sharing, reduce exploitation costs and improve travel experience, but it alsoblurs the lines between virtual world and reality with serious security matters. In the meanwhile airports face a dailychallenge to ensure business continuity and passengers’ safety.SATIE adopts a holistic approach about threat prevention, detection, response and mitigation in the airports, whileguaranteeing the protection of critical systems, sensitive data and passengers. Critical assets are usually protectedagainst individual physical or cyber threats, but not against complex scenarios combining both categories of threats.In order to handle it, SATIE develops an interoperable toolkit which improves cyber-physical correlations, forensicsinvestigations and dynamic impact assessment at airports. Having a shared situational awareness, security practitionersand airport managers collaborate more efficiently to the crisis resolution. Emergency procedures can be triggeredsimultaneously through an alerting system in order to reschedule airside/landside operations, notify first responders,cybersecurity and maintenance teams towards a fast recovery.Innovative solutions will be integrated on a simulation platform in order to improve their interoperability and tovalidate their efficiency. Three demonstrations will be conducted at different corners of Europe (Croatia, Italy andGreece) in order to evaluate the solutions in operational conditions (TRL≥7). Results and best practises will bewidely disseminated to the scientific community, standardization bodies, security stakeholders and the aeronauticcommunity. Finally, SATIE paves the way to a new generation of Security Operation Centre that will be included in acomprehensive airport security policy.

Automatic assessment of acute stress from speech.
Automatické hodnotenie akútneho stresu z reči
Annotation: A person often intuitively identifies whether his/her partner in conversation is under stress, just from stress manifestations in the voice and speech. However, the group of acute stress manifestations and their measurable acoustic cues are not known well. There are no suitable speech databases available to study these manifestations. We will design an experimental scenario when the subject is stressed and is forced to communicate intensely via speech. During the experiment other modes, such as video, biosignals (pulse, skin conductivity, …), will be recorded in addition to speech. We will analyze the collected data and propose an annotation scheme to capture the occurrence of stress. We will investigate the effects of stress and their acoustic correlates and select a set of acoustic features and a suitable machine learning method for automatically identifying stress in speech. A system for stress detection from speech will be developed and its results will be compared to subjective expert judgment.

Finished projects (annotations):

Grapheme and Phoneme Frequencies in Slovak
Frekvencie grafém a foném v slovenčine
Annotation: Research on the frequencies and propability densities of graphemes and phonemes in Slovak.

Global ATM security management (GAMMA)
Globálny bezpečnostný manažment riadenia letovej prevádzky
Annotation: The project deals with the global ATM security management in the European aerospace in future.

Intelligent Language Tutoring System with Multimodal Feedback Functions
Inteligentný systém pre výuku jazykov s multimodálnymi funkciami spätnej väzby
Annotation: Computer assisted foreign language tutoring using speech recognition methods.

Nonverbal speech gestures in Slovak in multimodal and multicultural context
Neverbálne rečové gestá v slovenčine v multimodálnom a multikulturálnom kontexte
Annotation: The project is aimed at the research of non-verbal speech gestures. The basic set of these gestures in current spoken Slovak is to be defined In the course of the project. This set will be investigated in relation to other modalities (text, lexical speech gestures, static video representation – picture/photo, and dynamic video representation – video/film.)The research will also include comparative studies in multi-cultural context. For the main goals we have chosen the research on these gestures in Slovak-Hungarian language pair (diametrically different languages, very close cultural and historical context), and Slovak-Japanese (diametrically different languages). The expected results will be both theoretical, with a new contribution to the knowledge on speech communication including sound speech gestures, and on their relation to text and video representations, as well as to the intercultural research in this area.The results will be applied practically in the expressive speech synthesis and the expressive speech recognition.

Automatic meeting speech recognition with application to courtroom recordings transcription.
Automatické rozpoznávanie reči pri komunikácii viacerých osôb s aplikáciou na prepis súdnych pojednávaní.
Annotation: Communication in multi-party meetings is one the main driving forces of progress in the current modern globalized society of extremely specialized fields. Reliable automated transcription of such meetings offers great potential for subsequent multi-layered analysis and data mining, and provides thus a promising source of information and knowledge. The proposed project deals with the automatic recognition and transcription of speech during multi-party meetings in the Slovak language using progressive methods and parallel computing. The core area of the project is the automatic recognition of speech in court proceedings.

Automatic transcription of dictate
Automatický prepis diktátu
Annotation: The aim of the project is the research and development of the automatic dictation transcription system for the Ministry of Justice of the Slovak Republic.

Automated, Voice Controlled Telecommunication Systems and their Applications
Automatizované, hlasom ovládané telekomunikačné systémy a ich aplikácie
Annotation: The project is focused on the applied research in the area of speech technologies for voice-operated telecomunication services with potencional influence in other areas of speech technologies, such as automatic speech transcription, search in the speech databases, automatic translation, data mining, semantic web etc.

Intelligent Language Tutoring System with Multimodal Feedback Functions
Inteligentný systém pre výuku jazykov s multimodálnymi funkciami spätnej väzby
Annotation: The project is aimed at co-financing the EACEA project Computer assisted foreign language tutoring using speech recognition methods (EURONOUNCE).

Nonverbal speech gestures in Slovak in multimodal and multicultural context
Neverbálne rečové gestá v slovenčine v multimodálnom a multikulturálnom kontexte
Annotation: This MVTS project co-funds the active participation of the researchers form the IISAV in the COST 2102 project. The research in this project is focused on non-verbal speech gestures.

Cross-linguistic study of prosodic cues in English, Spanish, and Slovak
Porovnávacia štúdia prozodických signálov v anglickom, španielskom a slovenskom jazyku
Annotation: The project deals with Cross-linguistic study of prosodic cues in English, Spanish, and Slovak.

Speech Technologies for Modern Telecommunication and Information Systems and Services
Rečové technológie pre moderné telekomunikačné a informačné systémy a služby
Annotation: The project is focused on the speech technologies for modern voice-operated telecommunication and information systems with potencional influence in other areas of speech technologies, such as automatic speech transcription, search in the speech databases, automatic translation, data mining, semantic web etc. The goals of the project are strictly focused on the Slovak language.

Robust Speech Technologies for Information Systems in Slovak and their Diagnostics
Robustné rečové technológie pre informačné systémy v slovenčine a ich diagnostika
Annotation: The main goal of the project is in major and measurable improvement of reliability and robustness of automatic speech processing systems in Slovak, and in improved communication naturalness of language interface in Slovak. In research of Slovak phonetics and phonology from the viewpoint of automatic processing, primarily defining of SAMPA, a more general phonetic alphabet (suitable for various application types) will be concerned. Within automatic speech recognition, the project aims at achieving higher reliability of speech recognizers for information systems in Slovak, and at development of diagnostic procedures to reveal speech recognizers drawbacks. Within speech synthesis, the goal of the project is in the development of models of individual speech prosody components in Slovak, in improved naturalness of synthetic speech, as well as in better intelligibility of synthetic speech even in environments with noise, and in the development of diagnostic procedures to reveal speech synthesizers drawbacks.

Speaker verification as an add-on element of the Air Traffic Management security
Verifikácia hovoriaceho ako prvok bezpečnosti riadenia letovej prevádzky
Annotation: The project is aimed at checking up the possibility of using the speaker verification (SV) as an add-on safety element in the ATM security. The objective of the project is to study the properties of this communication and to verify whether it would be possible to use the speaker verification system to verify the identity of the speakers and to unavil a potential entry by an intruder in this communication. In this project we want to focus on the initial assessment of the involvement of the speaker verification technology in the ATM security system. The SV and its resistance to fraud attacks will be performed in the Slovak languge too. A set of speech recording samples of 500 speakers will be prepared a unique set of 500 corresponding synthesizers with wide range of settings will be created, which together with the existing synthesizers will be used to prepare data for imitated attacks. The complex calculations related to training and evaluation will be realized on the computer cluster of the II SAS.