Projects and Grants We Are Involved In

Tools to Combat Voice Deepfakes (NABOSO)

Project Registration Number: VB02000060

Project Coordinator: Brno University of Technology, Faculty of Information Technology

Partners: Phonexia s.r.o.

Implementation: 01.01.2024 – 31.12.2026

Funding Body: Ministry of Interior of the Czech Republic

Annotation: This project focuses on understanding the possible uses and associated threats in the malicious use of voice deepfakes, with the intention of finding reliable strategies to mitigate these threats. By doing so, we can ensure that these innovations are used ethically and responsibly, creating a more sustainable and fairer future for all. Its activities are focused on three main areas: enhancing the robustness of voice biometric systems against voice deepfake spoofing attacks, creating tools to counter the malicious use of voice deepfakes, and evaluating people's ability to recognize voice deepfakes, while also finding ways to increase users' security awareness.

Support for participation of Phonexia s.r.o. at foreign trade fairs

Project Registration Number: CZ.01.02.01/04/23_017/0001524

Project Coordinator: Phonexia s.r.o.

Partners: no partners

Implementation: 19.06.2023 – 31.12.2025

Funding Body: Ministry of Industry and Trade of the Czech Republic; co-funded by the European Union

Annotation: The main goal of the project is to promote the company's overall product offering in order to strengthen its position in foreign markets and facilitate entry into these markets. The company will focus on markets in EU and NATO countries.

Reliable biometric technologies to assist police authorities in combating terrorism and organized crime - TENSOR

Project Registration Number: 101073920

Project Coordinator: Netcompany-Intrasoft (BE/LU)

Partners: Phonexia (CZ), Idemia Identity & Security Germany (DE), T3K-Forensics (AT), Fujitsu Services (DE), Eticas Research and Innovation (ES), Squaredev (BE), Quadible Greece (GR), ZITIS (DE), Ianus Consulting (CY), TechnischeUuniversiteit Delft (NL), Ethniko Kentro Erevnas Kai Technologikis Anaptyxis (GR), Idryma Technologias Kai Erevnas (GR), Inspectoratul General al Politiei (MD), Ministério da Justiça (PT), Policejní prezidium České republiky (CZ), Ministry of the Interior (FI), Ubitech Limited (CY)

Implementation: 01/2023 – 12/2025

Funding Body: European Union – The Horizon 2020 program

Annotation: The TENSOR project will provide LEAs a platform that facilitates biometric evidence extraction, sharing, and storage in cross-border environments allowing them to share best practices in an automated, robust, secure, privacy-preserving, and scalable manner. The full potential of biometric technologies will be exploited along with their fusion in order to strengthen their acceptance in court when partial evidence is gathered from the crime scenes.

This project acquired financial means from the Horizon 2020 project for research and innovation of the European Union program based on grant agreement no. 101073920.

Exchanges for SPEech ReseArch aNd TechnOlogies - ESPERANTO

Project Number: 101007666

Project Coordinator: LIUM (Laboratoire d'Informatique de l'Université de Mans)

Partners:

Academic Partners: ViVoLAB (Universidad Zaragoza); LNE (Laboratoire National de Métrologie et d'Essais); BUT Speech@FIT (Brno University of Technology); LIA (Laboratoire d'Informatique d'Avignon); LIG (Laboratoire Informatique de Grenoble); NLP Group (University of Sheffield); UNIMAS (Universiti Malaysia Sarawak); USM (Universiti Sains Malaysia); IDASCO (Informatique Distribuée Appliquée aux Systèmes Complexes); ICC (Insituto de Ciencias de la computación); LPTV (University of Chile); CENTATAV (Centro de Aplicaciones de tecnologias de Avanzada); JHU (Johns Hopkins University); MILA (Université de Montréal);
Non Academic Partners: Phonexia (CZ); Elyadata (TN); Omilia (EL); Allo-Media (FR);
Collaborating Institutions: NIST (National Institute of Standards and Technology).

Implementation: 1.1.2021 – 31.12.2025 (60 months) A group of yellow stars Description automatically generated with low confidence

Annotation: The project will contribute to the development of the next generation of AI; AI interacting with their users for better learning but also explainable systems: systems whose decisions are understandable by humans. In order to make these technologies accessible to the largest number of people, ESPERANTO will address applications with limited resources in the field of rare languages, robotics or education. ESPERANTO is a collaborative research program based on a partnership between 15 academic and 4 non-academic partners, having complementary expertise and resources.

This project has received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement No 101007666.

Real Time Network, Text, and Speaker Analytics for Combating Organized Crime – ROXANNE (Finished)

Project Registration Number: 833635

Project Coordinator: Idiap Research Institute
Partners: Phonexia s.r.o., Brno University of Technology, Police of the Czech Republic, Trïlateral Research Ltd., SAIL LABS Technology GmbH, Capgemini Consulting, The International Criminal Police Organization (INTERPOL), Saarland University, The Center for Security Studies (KEMEA), The University of Hanover, Catholic University of the Sacred Heart (Italy), AEGIS IT RESEARCH UG, Airbus Defence and Space SAS, Ministry of Internal Affairs (Romania), Forensic Science Centre of Lithuania (FSCL), Police Service of Northern Ireland, Advanced Integrated Technology Solutions & Services Ltd. (ADITESS), Ministry of the Interior (Croatia), Netherlands Forensic Institute, Internet Of Things Applications and Multilayer Development Ltd (IOTAM), Ministry of Public Security (Israel), Hellenic Police, Garda Síochána

Implementation: 09/2019 – 12/2022

Funding Body: European Union – The Horizon 2020 program

Annotation: The ROXANNE project’s goal is to create an effective tool for discovering organized crime and identifying persons of interest based on cutting-edge speech technologies and visual and link analysis. ROXANNE will significantly speed up investigation processes and improve the identification of persons of interest based on their voice in situations where the analysis of large amounts of legally obtained audio recordings (that can also be in different languages) is required. The project’s output will be a platform that is easy to use by law enforcement agencies for collecting evidence and other important information based on speech, language, and video technologies. The final version of the platform will use natural language processing (NLP) technology and relational analytics to visualize criminal activity clearly on a graph. ROXANNE will comply with all relevant legal and ethical regulations of the European Union and INTERPOL for sensitive data processing.

This project acquired financial means from the Horizon 2020 project for research and innovation of the European Union program based on grant agreement no. 833635.

Main project results: http://roxanne-euproject.org

Employment of artificial intelligence into an emergency call reception (Finished)

Project Coordinator: VSB - Technical University of Ostrava, Faculty of Safety Engineering
Partners: BornDigital s.r.o., Phonexia s.r.o., SpeechTech s.r.o., Brno University of Technology, Faculty of Information Technology

Implementation: 01/07/2019 – 31/05/2022
Funding Body: Ministry of the Interior of Czech Republic

Annotation: The projects investigates into deployment of artificial intelligence for the reception of emergency calls during crisis events with a voice chat-bot. We expect advancing speech analytics, semantic analysis, dialog management and speech synthesis, integrated with geographical information. The output is a working demonstrator with real telephone calls in conditions similar to deployment in the Integrated emergency system (IZS), re[1]commendations for integration and for further automation of IZS.

Speech Analysis System - STOPAŘ (Finished)

Project Registration Number: VH20202021050

Project Coordinator: Phonexia s.r.o.
Partners: no partners

Implementation: 05/2020 – 12/2021
Funding Body: Ministry of the Interior of Czech Republic – Czech Republic Security Research (2016 - 2021)

Annotation: The main aim is to support security systems that will facilitate the suppression of all forms of serious crime as much as possible. This requires a balanced system of prevention and repression, as well as the ability to monitor crime trends, respond to the use of new technologies or the misuse of digital tools for committing a crime, understand the principles of a criminal sphere adaptation to new demographic conditions, and map the rate of latent crime.

Main Project Results:

R01 (Software) “XL4_stopar2021_8khz” – an adapted model for the comparison of SID voiceprints and PyTorch-based software to use it.
R02 (Software) “CS_CZ_stopar2021_6” – an adapted acoustic model for S2T, an adapted language model for S2T, and KALDI-based software to use these.
R03 (Software) “Evaluation” – S2T and SID accuracy measurements on validation data sets.

Robotic Information Center Development (Finished)

Project Registration Number: CZ.01.1.02/0.0/0.0/19_262/0019917

Project Coordinator: Phonexia s.r.o.
Partners: ARTIN, spol. s.r.o.

Implementation: 10.1.2020 – 31.12.2021

Annotation: The project is focused on the development of a prototype of the Robotic Information Center (RIC) and its software parts by the applicant and the partner within the framework of effective cooperation. The RIC will find its application as a voice assistant in the public sector and customers’ business activities.

Information Mining From Speech Gathered By Far-Field Microphones – DRAPÁK (Finished)

Project Registration Number: VI20152020025

Project Coordinator: Brno University of Technology
Partners: Phonexia s.r.o.

Implementation: 10/2015 – 09/2020
Funding Body: Ministry of the Interior of Czech Republic – Czech Republic security research (2015-2020)

Annotation: Information mining from speech is becoming indispensable for agencies fighting crime and terrorism. Current versions allow successful deployment on signals acquired via “close-talk” microphones. The objective of the DRAPÁK project is to increase success rate of speech mining from far-field microphones in real environment and generate relevant information in adequate operation scenarios. The output is a set of software tools which will be available for testing to Czech Republic police forces and government agencies.

Trade Fairs and Exhibitions Abroad (Finished)

Project Registration Number: CZ.01.2.111/0.0/0.0/18_167/0012965

Project Coordinator: Phonexia s.r.o.
Partners: None

Implementation: 23.2.2018 – 31.10.2020

Annotation: The project is realized through the Operational Program for Business and Innovation Competitiveness (Operační program Podnikání a inovace pro konkurenceschopnost). The project aims to increase the competitiveness of Phonexia s.r.o. in foreign markets by supporting the company’s participation in 15 global fair trades and exhibitions focused on speech technologies. The project’s result will be the promotion of the company’s products and services and the entry into new markets.

The project is co-funded by the European Union.

Development of a Platform for the Unification of Speech Technology Products (Finished)

Project Registration Number: CZ.01.4.04/0.0/0.0/16_076/0008353

Project Coordinator: Phonexia s.r.o.
Partners: None

Implementation: 1.1.2017 – 31.12.2019

Annotation: The project is realized through the Operational Program for Business and Innovation Competitiveness (Operační program Podnikání a inovace pro konkurenceschopnost). The aim of the project is to create a software platform—Phonexia Speech Platform—that will unify all current speech technologies of the company into one product. The output of the project will contribute to gaining new customers, expanding the product portfolio, and increasing the company's competitiveness.

The project is co-funded by the European Union.

Big speech data analytics for contact centers - BISON (Finished)

Project Registration Number: 645323

Project Coordinator: Phonexia s.r.o.
Partners: Brno University of Technology, MyForce NV, Telefonica I+D, Ebos Luxembourg S.A., ComCzech a.s., Telefönica Móviles S.A., University of Bologna

Implementation: 01/2015 – 12/2017

Funding Body: European Union – Horizon 2020 program

Annotation: The BISON project focuses on important innovations in three different areas: 1) basic technologies of data mining from speech (a quickly adaptable system for new languages, domains and contact centers campaigns), 2) business results of data mining from speech (converted to KPI – Key Performance Indicators), 3) contact center support systems integrating speech mining and business results in a user friendly way. The output of the project will be two prototypes: 1) smallBison (end of the first year) shall be a functional system for real but partly limited deployment and feedback, 2) bigBison (end of the project) shall include full range of possibilities and will be fully integrated with hardware of contact centers and software infrastructure.

This project acquired financial means from the Horizon 2020 project for research and innovation of European Union program based on grant agreement No. 645323.

Meeting assistant – MINT (Finished)

Project Registration Number: TA04011311

Project Coordinator: Brno University of Technology
Partners: Lingea s.r.o., Phonexia s.r.o., Tovek s.r.o.

Implementation: 10/2014 – 12/2017

Funding Body: TA ČR, program: Alfa, 4th public contest

Annotation: The objective of the project is research and development in the area of audio processing during interactions between people (meetings, team discussions, meetings with customers, etc.) leading to creation of an intelligent assistant prototype for assistance during meetings (online), recordings processing from a meeting (offline) and with subsequent systematical storage and sharing of meetings recordings.

Project TA04011311 “Meeting assistant” is implemented with financial support of TA ČR.

Main project results: www.mint-project.cz

Development of Models for Automatic Speech Recognition Technology for the Arabic Language – Dialect "Gulf Arabic" (Finished)

Project Registration Number: CZ.01.1.02 / 0.0 / 0.0 / 16_045 / 0008519

Project Coordinator: Phonexia s.r.o.
Partners: None

Implementation: 1.2.2017 – 30.4.2017

Annotation: The project is realized through the Operational Program for Business and Innovation Competitiveness (Operační program Podnikání a inovace pro konkurenceschopnost). The subject of the project is the creation of models for Automatic Speech Recognition (ASR) for the Arabic language – dialect Gulf Arabic. Based on these requirements, a pronunciation dictionary and an acoustic and a language model will be created for the recognition of spontaneous conversational Arabic. The output of the project will contribute to expanding the product portfolio and increasing the company's competitiveness.

The project is co-funded by the European Union.

Social Semantic Emotion Analysis for Innovative Multilingual Big Data Analytics Market - MixedEmotions (Finished)

Project Registration Number: 644632

Project Coordinator: National University of Ireland, Galway
Partners: Phonexia s.r.o., Deutsche Welle, Brno University of Technology, Universidad Politécnica de Madrid, Universität Passau, Expert System, Paradigma Tecnológico SL, SindiceTech

Implementation: 04/2015 – 03/2017

Funding Body: European Union – Horizon 2020 program

Annotation: The MixedEmotions project develops innovative multi-modal applications for big data analysis which will analyze more complex emotional profile of user's behavior while using data from mixed input channels: multilingual sources of text data, A/V signal input (multilingual speech, audio, video), social media (social networks, comments), and structured data. Commercial applications (implemented as pilot projects) will be aimed on Social TV, brand reputation management and contact centers. To give sense to collected user interactions from various data sources, modalities, and languages is a big challenge which has not been fully explored in industrial context.

This project acquired financial means from the Horizon 2020 project for research and innovation of European Union program based on grant agreement No. 644632.

Main project results: www.mixedemotions-project.eu

Availability Of Automatic Speaker Verification To Wide Spectrum Of Users In Security Sector – ZAOM (Finished)

Project Registration Number: VG20132015129

Project Coordinator: Brno University of Technology
Partners: Phonexia s.r.o.

Implementation: 04/2013 – 09/2015

Funding Body: Ministry of the Interior of Czech Republic – Czech Republic security research (2010-2015)

Annotation: In last years, there has been a significant theoretical and algorithmic development in Speech Recognition Engine (SRE). The project focuses on algorithms adaptation for specific needs of police and intelligence agencies, so that they (1) provide precise but simple displaying of results, so that responsible personnel get required information in time and keep reaction speed against threats and in investigation, (2) to be able to perform additional training of these on target user data and significantly increase their success rate.

Project implementation proceeded according to the plan. Project was successfully finished on 30.09.2015.

Main project results:

Phonexia Speaker Identification (v2.5.1)
Phonexia Voice Inspector (v1.3.2)
Phonexia Speaker Identification (v3.3.1)

Please use the following code for downloading: 09DF8E8F35CF74B8.

Note: for 30 days trial version contact Phonexia team. License (license agreement) is always necessary.

Speech Processing For Effective Human – Computer Communication Technology – TŘPIT (Finished)

Project Registration Number: TA01011328

Project Coordinator: Brno University of Technology
Partners: Lingea s.r.o., Phonexia s.r.o., Optimsys s.r.o.

Implementation: 01/2011-12/2014

Funding Body: TA ČR, program: Alfa

Annotation: The goal of the project was to develop advanced techniques for speech recognition and deploy them in practical applications: searching in electronic vocabulary on mobile devices, translations dictation, in security and defense, in dialog systems, Customer Relationship Managements systems (CRM, helpdesk, etc.), and in audiovisual approach to learning materials.

Project implementation proceeded according to the plan. Project was successfully finished on 31.12.2014.

Project TA01011328 “Speech Processing For Effective Human – Computer Communication Technology” was implemented with financial support of TA ČR.