Czech-French AI Workshop hero image

Czech-French AI
Workshop on Artificial Intelligence

12 September 2022 | Ministry of Foreign Affairs
13 September 2022 | CIIRC CTU

organizing partners

logo Ministry of foreign affairs
logo French Embassy in Czech Republic
logo CIIRC CTU
logo Inria

with the support of

logo Impact
logo Elise
logo VISION
logo RICAIP
logos EU

About the Workshop

CZ-FR AI Workshop
Czech-French AI Workshop decorative image

The Czech-French AI Workshop is jointly organised by the Czech Ministry of Foreign Affairs, the French Embassy in Prague, the Czech Institute of Informatics, Robotics and Cybernetics CTU in Prague, and Inria, with the support of other partners and synergic networks.

The aim is to bring together leading experts from various AI fields, entrepreneurs and decision-makers. The event should be a platform for connecting the key stakeholders active in the development and application of digital technologies and AI. The workshop will include a high-level session with key stakeholders and a session on impact and industry.

The workshop is in-person or online upon registration – please register and join the discussions.

Strategic Background
Czech-French AI Workshop decorative image

AI is a potentially transformational technology that has broad social, economic, national security, and geopolitical implications. Therefore, AI applications could have wide-ranging economic impacts across many sectors including manufacturing, transportation, health, education, energy and climate.

EU countries need to catch up with the global leaders and expand their AI capacities in order to benefit from new digital technologies while preserving our basic values and interests such as individual integrity, human security, and economic prosperity.

Czech-French Collaboration in AI
Czech-French AI Workshop decorative image

France and the Czech Republic are ready to identify possible bilateral projects and support research exchanges and mobility that would add value to existing collaborations as in the EU-wide frameworks and those under consideration. They are determined to support each other in exploring the benefits of cross-border partnerships in research and AI technologies to be applied in the private sector.

Both countries are aware that the utilisation of AI is one of key challenges that would determine the future prosperity and geopolitical standing of the EU. The Czech-French collaboration should be open to partners and help build a resilient Europe, able to benefit from excellent and trustworthy AI.

Venue

The Czernin Palace (the headquarters of the Czech Ministry of Foreign Affairs), the CIIRC CTU and other places relevant to the Czech AI system (accessible through guided tours).

Online

Ask your questions on Sli.do: #CZFRAI

Program

The high-level segment of the event will be held on the first day and will be hosted by the Ministry of Foreign in the Czernin Palace.

The second day will be focused on specific scientific topics and will take place at the premises of the Czech Institute of Informatics, Robotics and Cybernetics Czech Technical University in Prague.

The scientific program is organized in five sessions each one being co-chaired by a Czech and a French researcher.

  • Ethics and social impact
  • NLP and Foundational models
  • Mathematics & optimization for AI
  • Robotics and embedded AI
  • Computer vision and trustworthy systems

12 September 2022 | Czernin Palace

  • 8:00 – 9:00
    Registration & Coffee
  • 9:00 – 9:30
    Welcome & Opening Remarks
    Moderator: Véronique Debord-Lazaro and Petr Kaiser
    • Jan Lipavský, Minister of Foreign Affairs of the Czech Republic
    • Alexis Dutertre, Ambassador of France 
    • Helena Langšádlová, Minister of Science, Research and Innovation of the Czech Republic 
    • Claire Giry, Director General for Research and Innovation, French Ministry of Higher Education, Research and Innovation
  • 09.30 – 11.00
    Keynotes and Panel Discussion: National AI Strategies & Priorities
    Moderator: Stephane Canu / Petr Kaiser
    • Renaud Vedel, French national coordinator on AI: 3 top priorities of the French AI Strategy: increasing and diversifying talents, embedded & frugal AI, trustworthy AI
    • Petr Očko, Deputy Minister, Czech Ministry of Industry and Trade: Czech innovation & IA strategies
    • Isabelle Herlin, Coordinator of the national research program on AI: Cooperation between Public & Private research in AI in the French context
    • Josef Šivic, Czech Institute of Informatics, Robotics and Cybernetics at Czech Technical University and AI Czechia
    • Cécile Huet, Head of unit Robotics and Artificial Intelligence Innovation and Excellence, DG CONNECT, European Commission (online)
  • 11:00 – 11:30
    Coffee Break
  • 11:30 – 13:00
    Panel Session on AI impact
    CZ-FR AI Projects | Collaborations | Challenges
    Moderator: Vladimír Mařík 
    • Patrick Pérez, Valeo.ai
    • Françoise Soulié-Fogelman, Hub France IA 
    • Christophe Leroux, CEA
    • Lukáš Kačena, prg.ai
    • Vít Dočkal, International Neurodegenerative Disorders Research Center
    • Alžběta Krausová, Institute of State & Law of the Czech Academy of Sciences
  • 13:00 – 14:00
    Lunch
  • 14:00 – 15:30
    Ethics and Social Impact 
    Chairs: Céline Castets-Renard and Emil Pelikán
    • Céline Castets-Renard, University of Ottawa, Faculty of Law
    • Monika Mareková
    • David Černý, Institute of State and Law of the CAS, Jiří Wiedermann, Institute of Computer Science of the CAS
  • 15:30 – 16:00
    Coffee break
  • 16:00 – 17:30
    NLP and Foundational models 
    Chair: Jan Hajič
    • François Yvon, LISN/CNRS
    • Jan Hajič, Faculty of Mathematics and Physics, Charles University
    • Benoit Sagot, Inria
    • Jan Černocký, Faculty of Information Technology, Brno University of Technology
  • 19:00
    Reception at the French Embassy (upon invitation)

13 September 2022 | CIIRC CTU

  • 9:00 – 10:30
    Mathematics and Optimization for AI
    Chairs: Didier Henrion and Jakub Mareček
    • Jérôme Bolte, Toulouse School of Economics
    • Alessandro Rudi, Inria
    • Jan Křetínský, Technical University of Munich
    • Thibault Gauthier, Czech Institute of Informatics, Robotics and Cybernetics at Czech Technical University
  • 10:30 – 11:00
    Coffee break
  • 11:00 – 12:30
    Robotics and embedded AI 
    Chairs: Nicolas Mansard and Robert Babuška
    • Nicolas Mansard, LAAS-CNRS 
    • Karla Štěpánová, Czech Institute of Informatics, Robotics and Cybernetics at Czech Technical University
    • Christian Wolf, NAVER LABS
    • Tomáš Svoboda, Faculty of Electrical Engineering at Czech Technical University
  • 12:30 – 13:30
    Lunch
  • 13:30 – 15:00
    Computer vision and trustworthy systems
    Chairs: Patrick Pérez and Josef Šivic
    • Cordelia Schmid, Inria
    • Jakub Mareček, Faculty of Electrical Engineering at Czech Technical University
    • Patrick Pérez, Valeo.ai
    • Josef Šivic, Czech Institute of Informatics, Robotics and Cybernetics at Czech Technical University and AI Czechia
  • 15:00 – 15:30
    Wrap-up & Closing Session
    Instruments to support Czech-French AI Cooperation, assessing what we have seen in the workshop, vision of the future.
    • Véronique Debord-Lazaro, French Embassy in the Czech Republic
    • Petr Kaiser, Ministry of Foreign Affairs of the Czech Republic
    • Vladimír Mařík, Czech Institute of Informatics, Robotics and Cybernetics at Czech Technical University 
    • Josef Šivic, Czech Institute of Informatics, Robotics and Cybernetics at Czech Technical University and AI Czechia
  • 15:30 – 16:30
    Guided Tour around Key AI Research Facilities I.
    CIIRC CTU: Testbed for Industry 4.0 | RICAIP Centre

Register


    Personal Information





    Workshop Events

    I register for:

    Speakers

    Jérôme Bolte
    Speaker photo

    Jérôme Bolte

    Toulouse School of Economics, ELLIS member Professor, Toulouse 1 Capitole University

    Research interests

    J. Bolte graduated in Montpellier in 2003 and was recruited at Université Pierre et Marie Curie, Paris, in 2004. He now holds an AI chair within the Toulouse institute ANITI and is affiliated with the University of Toulouse Capitole. He is a specialist in optimization, algorithms, and machine learning. He was awarded the SIAM Optimization prize 2017 for his collaborative work on first-order algorithms in nonconvex optimization.

    A Glance at Nonsmooth Automatic differentiation

    I will recall the fundamental role of nonsmooth automatic differentiation as the core learning mechanism in modern AI. I will then show how a recent theory we developed — «Conservative gradients» — helps to understand fundamental phenomena, such as the convergence of learning phases in deep learning, the optimization of learning parameters, the nonsmooth cheap gradient principle, or the differentiation of algorithms (solvers).

    Céline Castets-Renard
    Speaker photo

    Céline Castets-Renard

    University of Toulouse

    Full-time Professor of Law at the Toulouse Capitole University and Junior Member of the IUF since 2015. She is a director assistant of the IRDEIC, Jean Monnet Center of Excellence. She is a member of the Toulouse Capitole University Board since 2016 and of the Faculty of Law and Political Sciences Council since 2014. She is a member of the Scientific Committee of the Chaire L.R. Wilson. She is also a member collaborator of the Group “Law, Change and Governance”, which associates the Universities of Montreal, Mc Gill and Laval. She has created a Master’s degree in right of Digital Law at the University Toulouse Capitol which she manages. Her research project led within the framework of the IUF is on “Normative regulation of the digital world”.

    From Ethic to AI Law: what are the main provisions of the AI Act regarding social issues?

    Jan Černocký
    Speaker photo

    Jan Černocký

    Faculty of Information Technology, Brno University of Technology
    Head of Dept. of Computer Graphics and Multimedia

    Jan Černocký is an academic researcher and Associate Professor at FIT BUT. He founded the BUT Speech@FIT research group in 1997 and serves as its executive director. He graduated from BUT (Ing.) and from Université Paris Sud (PhD) and was with ESIEE Paris, France and OGI Portland, Oregon, USA. His research interests include artificial intelligence, signal processing and speech data mining (speech, speaker and language recognition). He is a Senior Member of IEEE and a member of the International Speech Communication Association (ISCA). He co-founded Phonexia, which is currently one of the world’s most important players in production speech technologies.

    David Černý

    David Černý

    Institute of State and Law, CAS

    Why machine consciousness needs ethics
    (in partnership with J. Widermann)

    Vít Dočkal
    Speaker photo

    Vít Dočkal

    INDRC – International Neurodegenerative Disorders Research Center
    Director

    Vít Dočkal graduated in 2006 from the Faculty of Social Studies, Masaryk University with two doctoral degrees from the Department of Political Science, and International Relations and European Studies.
    In the years 2009 – 2013 he had been leading the ICRC Project Management Office at St. Anne’s University Hospital in Brno – a large R&D infrastructure funded by € 180M from the European Structural and Investment Funds (ESIF) and the state funds. Since 2013 he works as the professional project manager at the Czech Institute of Informatics, Robotics, and Cybernetics of the Czech Technical University in Prague (CIIRC CTU). As the head of the Project Management Office (2013 – 2021), he was responsible for the strategic project management agenda of € 100M+.
    Vít Dočkal is closely involved in CLAIRE’s engagement with industry and in supporting the AI ecosystem in Central and Eastern Europe. He is also co-founder of the start-up company TRIX Connections. Together with Martin Tolar, he stands behind the idea of INDRC Centre which he is currently leading as its director.

    Alexis Dutertre

    Alexis Dutertre

    Embassy of France
    Ambassador of France to the Czech republic

    Why machine consciousness needs ethics
    (in partnership with J. Widermann)

    Thibault Gauthier
    Speaker photo

    Thibault Gauthier

    Czech Technical University, CIIRC

    After defending my PhD at the University of Innsbruck on the subject of “Learning-Assisted Reasoning within Proof Assistants” in 2018, I became a postdoctoral researcher working at Czech Technical University in Prague. I am the main developer of the tactic-based theorem prover TacticToe and the conjecture-making system QSynt.

    Towards a Fully Automated Proof Assistant

    A proof assistant is an interactive program that helps a user prove a conjecture from a trusted logical kernel. Proof assistants have successfully been used to verify proofs of famous theorems such as the Kepler conjecture or the 4-colour theorem. However, for this verification to work, many intermediate proof steps have to be given to the proof assistant. If the gap between reasoning steps is too large the proof assistant will fail to check the proof and ask the user for more intermediate steps. In this talk, we present proof automation techniques currently being developed to bridge larger and larger reasoning gaps.

    Claire Giry

    Claire Giry

    French Ministry of Higher Education and Research
    Director General for Research and Innovation

    Jan Hajič

    Jan Hajič

    Charles University

    Jan Hajič is a full professor of Computational Linguistics at the Institute of Formal and Applied Linguistics at the School of Computer Science, Charles University in Prague, where he has also received his PhD in 1995. He served as the head and deputy head of the Institute since 2001. His interests cover morphology and part-of-speech tagging of inflective languages, machine translation, deep language understanding, and the application of statistical methods in natural language processing in general. He also has extensive experience in building language resources for multiple languages with rich linguistic annotation and is currently the director of a large, multi-institutional research infrastructure on language resources in the Czech Republic, LINDAT/CLARIAH-CZ, which aims at making datasets and corpora openly available for linguistic and Digital Humanities research. His work experience includes both industrial research (IBM Research Yorktown Heights, NY, USA, 1991-1993) and academia (Charles University in Prague, Czech Republic and Johns Hopkins University, Baltimore, MD, USA, 1999-2000, adjunct position at University of Colorado, USA, 2017-2022). He has published more than 200 conference and journal papers, a book on computational morphology, and several other book chapters, encyclopedia and handbook entries. He regularly teaches basic and advanced courses on Statistical NLP and has multiple experiences giving tutorials and lectures at various international training schools. He has been the PI or Co-PI of numerous international as well as large national grants and projects (including EU Framework Programme projects, such as H2020 and HE, and the NSF ITR program in the U.S.). He is the chair of the Executive Board of META-NET, a European research network in Language Technology.

    Towards Natural Language Understanding

    Didier Henrion

    Didier Henrion

    Université de Toulouse

    Cécile Huet
    Speaker photo

    Cécile Huet

    DG Connect, European Commission
    Head of unit Robotics and Artificial Intelligence Innovation and Excellence

    Cécile Huet is Head of the Unit “Robotics and Artificial Intelligence Innovation and Excellence” at the European Commission. This unit funds and assists beneficial robotics and AI developments within Europe. Under Horizon Europe, the unit launched the Public-Private Partnership on AI, Data and Robotics This unit is also at the heart of the Coordinated Plan on Artificial Intelligence, and the European ‘ecosystem of excellence’ in AI. Cécile joined the unit since its creation in 2004. Previously, she worked for the industry in signal processing after a post-doc at the University of California Santa Barbara and a PhD at University of Nice Sophia Antipolis, France. In 2015, she has been selected as one of the “25 women in robotics you need to know about”.

    Strengthening the European AI Ecosystem of Excellence

    Lukáš Kačena
    Speaker photo

    Lukáš Kačena

    prg.ai

    An experienced consultant specializing in public administration focused on strategic management and planning, analysis, as well as research, development and innovation policy. Since April 2022 in the role of prg.ai Managing Director, with the aim to transform Prague into a significant European AI hub.

    Alžběta Krausová

    Alžběta Krausová

    Institute of State and Law of the Czech Academy of Sciences

    Jan Křetínský

    Jan Křetínský

    Technical University of Munich

    Helena Langšádlová

    Helena Langšádlová

    Minister of science, research and innovation

    Christophe Leroux
    Speaker photo

    Christophe Leroux

    CEA

    Christophe Leroux received a PhD degree in Decision Theory from Sorbonne University in Paris in 1990. She is a senior researcher in AI for robotics. He is currently Manager of European Affairs at the LIST institute from CEA Paris-Saclay. He is in particular in charge of the development of Smart Manufacturing activities. He represents CEA in research and innovation projects on AI and robotics in different sectors of activity. He coordinates the European project RIMA aiming at the creation of a network of Digital Innovation Hubs for Inspection and Maintenance of Infrastructures. He is also involved in other creation of Digital Innovation Hubs networks in Healthcare and Manufacturing to favour uptake of AI and robotics technologies by the European Industry. Christophe Leroux is a member of the Board of Directors of the euRobotics association where he represents CEA. euRobotics gathers robotics industry and research organizations. It supports the elaboration and supervision of the robotics part of the AI, Big data and Robotics European research programme. euRobotics provides support to the ADRA association and the European Commission in close collaboration with the AI and Big Data associations. He is a scientific advisor of the APPROCHE association aiming at developing rehabilitation and assistive robotics and AI solutions for handicapped people. Christophe Leroux is the author of several scientific papers and reports on human-robot interaction in International conferences journals and communications, on ethical, legal and socio-economic issues in robotics and on support for innovation in robotics.

    Jan Lipavský

    Jan Lipavský

    Ministry of Foreign Affairs of the Czech Republic

    Nicolas Mansard
    Speaker photo

    Nicolas Mansard

    Université de Toulouse

    Nicolas Mansard is a researcher at LAAS-CNRS in the Gepetto team. He is working with humanoid robots and other complex machines with legs and arms, that he would very much like to see evolve in the real world. For that, he and his team are developing transversal methodologies, hybridizing advanced numerical optimization with off-line motion learning, to produce successful experiments on real robots. He is the coordinator of the EU project “Memory of Motion” project which is just about to finish, and of the EU project “Agimus” in collaboration with CTU-Prague, INRIA-Paris, PAL-Robotics, Airbus and several other industrial partners.

    Learning and optimizing dynamic robot movements with whole-body predictive control and memory of motion

    Jakub Mareček

    Jakub Mareček

    Czech Technical University, FEE

    Jakub Marecek received a PhD degree in computer science from the University of Nottingham, Nottingham, U.K., in 2012. He has been a faculty member at the Czech Technical University in Prague, the Czech Republic, since 2020. Previously, he had also worked in two start-ups, at ARM Ltd., at the University of Edinburgh, at the University of Toronto, at IBM Research – Ireland, and at the University of California, Los Angeles. His research interests include the design and analysis of algorithms for optimisation and control problems across a range of application domains. He has co-authored more than 17 journals and 40 conference papers. He is a co-inventor on 9 families of awarded U.S. patents and several pending patent applications. Dr Marecek has been a member of the Mathematical Optimization Society since 2009. He has received a variety of awards for his work, including IBM’s Outstanding Technical Achievement Award, in 2019, and the ITS Ireland’s 5th Annual Award for Outstanding Public Project, in 2016. 

    Human-Compatible Artificial Intelligence with Guarantees

    In machine learning and artificial intelligence (AI), training data often capture the behaviour of multiple subgroups of some underlying human population. Recent regulations in both Europe and the US stipulate a number of requirements on the use of AI in such applications, which are centred on the notions of equal treatment and equal impact. In this talk, we will introduce the regulation, illustrate our work towards implementing the rather abstract notions, and briefly mention further work planned along these directions within Horizon Europe project “Human-Compatible Artificial Intelligence with Guarantees” (project id 101070568).

    Monika Mareková

    Monika Mareková

    Institute

    Petr Očko

    Petr Očko

    Deputy Minister, Czech Ministry of Industry and Trade

    Emil Pelikán

    Emil Pelikán

    Institute of Computer Science, CAS

    Patrick Pérez
    Speaker photo

    Patric Pérez

    VALEO.AI, ELLIS member, member of the ELISE Industrial Advisory Board

    Patrick Pérez is Valeo’s VP of AI and Scientific Director of valeo.ai, an AI research lab focused on Valeo automotive applications, self-driving cars in particular. Before joining Valeo, Patrick Pérez was a researcher at Technicolor (2009-2018), Inria (1993-2000, 2004-2009) and Microsoft Research Cambridge (2000-2004). His research interests include multimodal scene understanding and computational imaging.

    Safer driving AI with limited supervision

    Alessandro Rudi
    Speaker photo

    Alessandro Rudi

    INRIA, ELLIS member

    Machine learning is spreading in many fields, despite its huge needs in computing. To face these problems, Alessandro Rudi (Inria of Paris) received an ERC Starting Grant and set up the REAL project within the Sierra project team.

    Benoît Sagot
    Speaker photo

    Benoît Sagot

    INRIA (ELISE Associate Partner)

    Benoît Sagot, Directeur de Recherches (Senior Researcher) at Inria, is the head of the Inria project-team ALMAnaCH in Paris, France. A specialist in natural language processing (NLP) and computational linguistics, his research focuses on language modelling, language resource development, machine translation, text simplification, part-of-speech tagging and parsing, computational morphology and, more recently, digital humanities (computational historical linguistics and historical language processing). He has been the PI or co-PI of a number of national and international projects and is the holder of a chair in the PRAIRIE institute dedicated to research in artificial intelligence. He is also the co-founder of two start-ups where he uses his expertise in NLP and data mining for the automatic analysis of employee survey results.

    Large-scale language models and their training corpora

    Large-scale language models have received a lot of attention in the last few years, both because of how they allow improving the state of the art in most natural language processing (NLP) tasks and because of the scale of computational resources and training data they require to be trained. Collecting such training data is a highly non-trivial task and has a significant impact on the behaviour of the models trained on it, depending for instance on the data’s homogeneity and size. Moreover, the way such data is fed to the neural architecture, namely the way it is tokenised, also has an impact, especially in multilingual settings. In this talk, I will speak about how we developed the large-scale multilingual OSCAR corpus. I will describe the lessons we learned while training the French language model CamemBERT, the first large-scale monolingual model for a language other than English, especially in terms of the influence of size and heterogeneity of the training corpus. I will also sketch out a few research questions related to biases in large-scale language models, with a focus on the impact of tokenisation. I will conclude with my thoughts on the future of language models and their impact on NLP and other data processing fields (speech, vision).

    Tereza Šamanová

    Tereza Šamanová

    _

    Cordelia Schmid
    Speaker photo

    Cordelia Schmid

    INRIA, ELLIS Fellow INRIA Research Director, Head of the THOTH project-team

    Cordelia Schmid holds a M.S. degree in Computer Science from the University of Karlsruhe and a Doctorate in Computer Science, from the Institut National Polytechnique de Grenoble (INPG). Her doctoral thesis on “Local Greyvalue Invariants for Image Matching and Retrieval” received the best thesis award from INPG in 1996. She received the Habilitation degree in 2001 for her thesis entitled “From Image Matching to Learning Visual Models”. Dr. Schmid was a post-doctoral research assistant in the Robotics Research Group of Oxford University in 1996–1997. Since 1997 she has held a permanent research position at Inria, where she is a research director. Dr. Schmid is a member of the German National Academy of Sciences, Leopoldina and a fellow of IEEE and the ELLIS society. She was awarded the Longuet-Higgins prize in 2006, 2014 and 2016 and the Koenderink prize in 2018, both for fundamental contributions to computer vision that have withstood the test of time. She received an ERC advanced grant in 2013, the Humbolt research award in 2015, the Inria & French Academy of Science Grand Prix in 2016, the Royal Society Milner award in 2020 and the PAMI distinguished researcher award in 2021. Dr. Schmid has been an Associate Editor for IEEE PAMI (2001–2005) and for IJCV (2004–2012), editor-in-chief for IJCV (2013–2018), a program chair of IEEE CVPR 2005 and ECCV 2012 as well as a general chair of IEEE CVPR 2015, ECCV 2020 and ICCV 2023. Starting 2018 she holds a joint appointment with Google research.

    Do you see what I see? Large-scale learning from multimodal videos

    In this talk, we present recent progress on large-scale learning of multimodal video representations. We start by presenting VideoBert, a joint model for video and language, repurposing the Bert model for multimodal data. This model achieves state-of-the-art results on zero shot prediction and video captioning. Next, we present an approach for video question answering which relies on training from instruction videos and cross-modal supervision with a textual question answer module. We show state-of-the-art results for video question answering without any supervision (zero-shot VQA) and demonstrate that our approach obtains competitive results for pre-training and then fine-tuning on video question answering datasets. We conclude our talk by presenting the recent VideoCC dataset, which transfers image captions to video and allows obtaining state-of-the-art performance for zero-shot video and audio retrieval and video captioning.

    Françoise Soulié-Fogelman
    Speaker photo

    Françoise Soulié-Fogelman

    Hub France IA

    Françoise Soulié Fogelman has over 40 years’ experience in AI (neural networks), machine learning, social network analysis and big data both in academia and industry. A former graduate from École Normale Supérieure, she holds a PhD from University of Grenoble. She was Professor at the University of Paris 11-Orsay, where she was advisor to 20 PhDs (neural networks, deep learning). She then funded a startup (Mimetics to develop a neural-network based OCR product) to later join Atos (as head of a data mining – data warehouse group) and Business & Decision (as Partner) where she created and headed the CRM business unit. At KXEN, she was Vice President Innovation until the company was bought out by SAP. She then joined the School of Computer Software at Tianjin University (China), where she was a professor, head of the Data Science team. She is now Scientific Advisor for Hub France IA. She has co-authored more than 150 scientific publications and 13 books. She is/has been an expert for the European Commission, ANR (French National Research Agency), French Competitivity cluster Cap Digital and CCF Big Data Task Force (China). She is a founding member and member of the board of Hub France IA (read AI). She was a member of the AI High Level Experts group for the European Commission and is co-chair of the working group “innovation & commercialization” for the GPAI (Global Partnership on AI).

    Tomáš Svoboda
    Speaker photo

    Tomáš Svoboda

    Czech Technical University in Prague, Faculty of Electrical Engineering
    Dept. of Cybernetics

    Tomáš Svoboda is a professor at the Czech Technical University in Prague and he is the Chair of the Department of Cybernetics. Before joining the department in 2003, he spent three postdoctoral years with the Computer Vision Group at ETH Zurich, Switzerland. He serves as the Director of the Ph.D. program Cybernetics and Robotics. He has published papers on multi-camera systems, omnidirectional cameras, image-based retrieval, learnable detection methods, robot learning, and USAR robotics; He led the successful CTU-CRAS-NORLAB multi-robot team at the DARPA SubTerranean Challenge (2018-2021). His recent research interests include multimodal perception for autonomous systems, object detection, and related applications in autonomous systems including automotive.

    Robots go deep – multi-robot missions in unknown undergrounds

    Renaud Vedel
    Speaker photo

    Renaud Vedel

    French national coordinator on AI

    Renaud VEDEL joined the French Civil Service 20 years ago. He has been serving as Prefect since 2012. He served as junior State Local Representative, budgetary and legal advisor to the director general of the national police, secretary general of the police prefecture of Paris. He served the Government as deputy-director of the office of the Minister of the Interior, then as senior adviser to the Prime Minister on Home affairs, security and intelligence. Since 2018, he is committed to France’s Digital policies and its National AI Strategy. First, as coordinator at ministerial level (Ministry of the Interior). Since March 2020, by mission letter from the Prime Minister, he coordinates the National Strategy for AI.

    3 top priorities of the French AI Strategy: increasing and diversify talents, embedded & frugal AI, trustworthy AI

    Jiří Wiedermann
    Speaker photo

    Jiří Wiedermann

    Institute of Computer Science of the CAS

    Jiří Wiedermann belongs to the first generation of computer scientists graduating in former Czechoslovakia. He has worked in informatics since the beginning of his scientific carrier in nineteen seventies. His recent research interests are in the theory of artificial intelligence (machine consciousness). By his invited talks, publications and activities in organization of important European (Mathematical Foundations Of Computer Science – MFCS, International Colloquium on Algorithms and Languages  – ICALP) and national computer science conferences (Software Seminar  – SOFSEM) he contributed to the development of informatics both at national and international level. In nineteen nineties he acted as the vice-president of the European Association for Theoretical Computer Science (EATCS). Between 2000 and 2012 he served as the director of Institute of Computer Science of Czech Academy of Sciences (CAS), at present he is the member of the Scientific Council of CAS. Professor Wiedermann was a member of the board of directors of ERCIM (European Research Consortium in Informatics and Mathematics) (1997-2010) and is a member of Academia Europaea (London) and of the Czech Learned Society (Prague).

    Machine consciousness: the road to better AI (in partnership with David Cerny)

    Current research into AI aligned with human rights and values focuses on the notion of trustworthy AI (T-AI). Alas, considerably less attention has been paid to machine consciousness, which we consider an essential component of the concept of T-AI. It requires that T-AI be fully aware of both the states of its parts and the surrounding environment. We present the main results of our research on machine consciousness with its ethical presuppositions and implications and show that endowing AI-based systems with machine consciousness demonstrably leads to more trustworthy and safer AI. We believe that machine consciousness represents a crucial step toward achieving the goal of developing human-centered AI.

    Christian Wolf
    Speaker photo

    Christian Wolf

    Naver Labs Europe

    Christian WOLF is Principal Scientist at Naver Labs Europe. He is interested in AI for Robotics, in particular machine learning and embodied computer vision; large-scale learning of the capacity to perform high-level reasoning from visual observations, and more recently the connections between machine learning and control. He is a member of the directing committee of GDR ISIS and co-leader of it’s topic “Machine Learning”; member of the scientific committee of GDR IA; member of the board of AI experts at the French national supercomputing cluster GENCI; member of evaluation ANR committee “Artificial Intelligence” from 2019-2021. He has supervised 15 defended PhD theses, is an associate editor of IEEE-Transactions on PAMI and area chair of CVPR 2020, NeurIPS 2020, ICLR 2021, ICCV 2021, ICML 2021, NeurIPS 2021, ICML 2022, ECCV 2022. From 2005 to 2021 he was associate professor (Maître de Conférences, HDR) at INSA de Lyon and LIRIS, a CNRS laboratory, where he was also the head of the AI chair / chair in Artificial Intelligence (the group). He received his MSc in computer science from TU Vienna, Austria, in 2000, and a PhD in computer science from INSA de Lyon, France, in 2003. In 2012 he obtained the habilitation diploma, also from INSA de Lyon.

    Learning representations for visual navigation in 3D environments

    In this talk we address perception and navigation problems in robotics settings, in particular mobile terrestrial robots and intelligent vehicles acting in 3D environments from visual input. We focus on learning representations, which are structured and allow to reason on a high level on the presents of objects and actors in a scene and to take planification and control decisions. In particular, we compare different ways to design inductive biases for deep reinforcement learning: neural metric maps, neural topological maps, and neural implicit representations.

    François Yvon
    Speaker photo

    François Yvon

    LIMSI (CNRS / Université Paris Saclay)

    François Yvon is a senior CNRS researcher at the LISN (formerly LIMSI) laboratory of Université Paris Saclay in Orsay, France. F. Yvon has been leading activities in Machine Translation at LISN for about 15 years, resulting in more than one hundred scientific publications on all aspects related to the development and evaluation of multilingual language processing technologies, from word and sentence alignment to translation modelling and evaluation, including recent work on multi-domain adaptation in MT and on cross-lingual transfert learning issues. He has acted as coordinator or PI in multiple past national and international projects in MT such as Quaero or H2020/QT21 and has supervised more than 15 PhDs on MT related topics. Between 2013 and 2020, Dr. Yvon has also been the general director of the LIMSI laboratory. He is a board member of the European chapter of the Association for Computational Linguistics, of the MetaNet network, and has recently contributed as an expert on linguistic technologies for the French language to several European projects (European Language Resource Collection, ELE – European Language Equality, ELG – European Language Grid).

    How multilingual are Large Multilingual Language Models ?

    The recent developments of very large language models (LLMs) has suggested that, when trained with a mixed language set of data, these models could learn multilingual representations useful in transfert learning scenarios, and could also perform multilingual natural language processing task, such as Machine Translation. In this talk I will discuss and may be challenge the multilingual capacities of these large LMs, based on a serie of experiments performed with openly accessible versions of these models, such as mBart, mT5, mGPT and BLOOM.

    Karla Štěpánová
    Speaker photo

    Karla Štěpánová

    CIIRC CTU

    Karla Stepanova is a researcher and a team leader at the Czech Institute of Informatics, Robotics, and Cybernetics in Prague (CIIRC CTU). She received her Ph.D. degree in Artificial Intelligence and Biocybernetics at FEE CTU in Prague, in 2017. She spent 6 months at Plymouth University in Angelo Cangelosi’s lab working with the iCub robot and currently is focused on building Imitation Learning Centre at CIIRC CTU. Her research interests are probabilistic models of cognition, unsupervised learning, multimodal integration, language acquisition, symbol grounding, and learning by demonstration.
    Personal webpage: http://karlastepanova.cz/
    Team webpage: http://imitrob.ciirc.cvut.cz

    Verbal and non-verbal communication with robots to teach them new skills


    Organising committee

    Véronique Debord-Lazaro
    Organiser photo
    logo

    Véronique Debord-Lazaro

    Scientific and Higher Education Attachée, French Embassy in the Czech Republic

    Ms Debord-Lazaro graduated from the Institute of Political Science in Bordeaux and later earned a master’s degree in geography from the University of Poitiers. She worked for many years at French universities in the field of international affairs and European projects; prior to joining the French Embassy in the Czech Republic as a science attaché, she was director of the international office of the University of Bordeaux for several years. In her current capacity, she supports Czech-French cooperation in education and research, in particular thanks to several support programmes (visit the IFP website).

    Petr Kaiser
    Organiser photo
    logo

    Petr Kaiser

    Special Envoy for Science Diplomacy, Ministry of Foreign Affairs of the Czech Republic

    As a career diplomat with many years of experience, Mr. Kaiser has been dealing with scientific diplomacy. For many years, he was involved in multilateral diplomacy such as a mission to the UN in New York in the position of Deputy Permanent Representative and previously a mission to international organizations based in Vienna. He also worked at the National Gallery in Prague as an executive director and director of the science section. In the early 1990s, Mr. Kaiser worked as a researcher at the Academy of Sciences of the Czech Republic.

    Vladimír Mařík
    Organiser photo
    logo

    Prof. Vladimír Mařík

    Founder and Scientific Director of the CIIRC CTU

    Prof. Vladimíř Mařík acted as the managing director of the Rockwell Automation Research Center, Prague and also the Head of the Department of Cybernetics at CTU in 1999-2013 and a founder and Scientific Director of the CIIRC of the CTU in Prague. He has been a member of the R&D&I Council of the CZ (2011-15 and 2018-). He is co-author of the National Initiative for Industry 4.0 of the CZ and a Board member of AICZECHIA. He is co-author, principal investigator and coordinator of RICAIP. He acted as coordinator of the National Competence Centre on Cybernetics and AI (TA CR, EUR 8M). He acted as Editor-in-chief of IEEE SMC Part C (2005-13), Board member of IEEE SMC (2005 to present), has received the IEEE SMC Outstanding Contribution Award (2012) and has been Vice-president of IEEE SMC (2013-).

    Josef Šivic
    Organiser photo
    logo

    Josef Šivic

    Distinguished researcher at Czech Institute of Informatics, Robotics and Cybernetics at Czech Technical University, Chairman of AI Czechia, Director of ELLIS Unit Prague

    Josef Sivic holds a distinguished researcher position at the Czech Institute of Robotics, Informatics and Cybernetics at the Czech Technical University in Prague where he heads the Intelligent Machine Perception project and the ELLIS Unit Prague. He is currently on leave from a senior researcher position at Inria Paris where he remains a close external collaborator of the Willow team. He received a habilitation degree from Ecole Normale Superieure in Paris in 2014 and PhD from the University of Oxford in 2006. After the PhD, he was a post-doctoral associate at the Computer Science and Artificial Intelligence Laboratory at the Massachusetts Institute of Technology. He received the British Machine Vision Association Sullivan Thesis Prize, three test-of-time awards at major computer vision conferences, and an ERC Starting Grant. 

    Day 1:
    Czech AI research, impact and the European perspective

    Day 2:
    Learning embodied perception from instructional videos    

    Isabelle Herlin
    Organiser photo
    logo

    Isabelle Herlin

    Coordinator of the national research programme on AI, INRIA

    Isabelle Herlin is a former student of the Ecole Normale Supérieure de Sèvres. She holds a PhD in computer science and the Habilitation to Supervise Research (HDR) in mathematics from University Paris 6. Isabelle Herlin began her career as a research professor at University Paris 6 before being recruited as a research fellow at Inria. Research Director since 1996, Isabelle Herlin has successively created two Inria research teams, in image processing and data assimilation. Isabelle Herlin is coordinator of the National AI Research Program and director of the Paris center of expertise of the Global Partnership in Artificial Intelligence (GPAI). This center of expertise supports two working groups on the future of work and Innovation and Commercialization.

    Cooperation between public & private research in IA

    Contact

    If you have any questions about the organisational issues, program or participation, please do not hesitate to contact us.

    Contact Person photo

    Eva Doležalová
    Marketing Manager, CIIRC CTU eva.dolezalova@cvut.cz