International Research Projects (IRP)

The Intelligent Systems Diagnostics and Control Approaches IRP (ADONIS), 2020-2025, focuses on intelligent systems diagnostics and control. It brings together researchers from four partner organizations: Compiègne University of Technology (UTC), Faculty of Engineering – Lebanese University (UL), CNRS France and CNRS Lebanon, with common interests and a willingness to collaborate in the areas of control, data analysis, control of uncertainties and this in several frameworks of studies, such as in particular biomedical systems and transport systems. Three UTC/CNRS research units are involved in this IRP: Heuristics laboratory and diagnosis of complex systems (Heudiasyc - CNRS/Université de technologie de Compiègne), Roberval laboratory - Mechanical, acoustic and materials research unit (Roberval - CNRS/Université de technologie de Compiègne) and the Biomechanics and Bioengineering Laboratory (BMBI - CNRS/Université de technologie de Compiègne).

After many years of collaboration between these institutions, and particularly between UTC and UL since 1997, this project aims to consolidate and sustain this collaboration, to broaden its scope to new research themes, and increase its attractiveness and visibility.

See also: "Sustain and amplify Franco-Lebanese scientific collaboration"

One of the major current challenges in child-robot interaction within an educational framework is to enable effective and beneficial interactive learning over time. The robot must adapt online to different children and their progress. In return, the child should advance in their learning through interaction with the robot. The aim of this IRP, led by the Institut des systèmes intelligents et de robotique  (ISIR – CNRS/Sorbonne University), is to strengthen a partnership with the Polytechnic University of Athens. In recent years, ISIR researchers have pioneered the implementation of online learning capabilities in humanoid robots during interactions with typically developing children or those with autism spectrum disorders. The goal now is to demonstrate that this provides a significant long-term educational benefit compared to pre-programmed robots.

The IRP Game Theory for Market Design, AI and Fairness (FAIRGAME) aims to develop tools from game theory and social choice theory to design fair mechanisms in markets, AI and collective systems. Established in January 2026, it involves units from the Laboratoire d'analyse et de modélisation de systèmes pour l'aide à la décision (LAMSADE - CNRS/Université Paris Dauphine - PSL) and the Moroccan Centre for Game Theory (MCGT, UM6P).

The aim of the FAIRGAME IRP is to establish an international research hub in game theory at the interface of mathematics, economics and computer science (mechanism and market design, multi-agent learning, fairness and robustness in AI). Throughout its activities, the IRP organises academic exchanges, joint thesis supervision, thematic schools and international seminars.

Project leaders: Jérôme Lang, CNRS Research Director at LAMSADE, and Rida Laraki, CNRS Research Director seconded to MCGT/UM6P.


Key participants: O. Gorelkina, A. Kushnir, B. Missaoui, O. Saadi, N. Touzi, A. Abbou; P. Cardaliaguet, G. Carlier, Y. Chevaleyre, L. Gourvès, S. Moretti, D. Peters, R. Sanver, G. Vigeral, B. Ziliotto, J. Renault; J. Hartline, E. Elkind, J. Correa, P. Mertikopoulos.

The Geometric Deep Learning and Generative Models for 3D Human IRP (GeoGen3DHuman) between Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL - CNRS/Université de Lille/ Centrale Lille) and the Media Integration and Communication Center (MICC) is a joint research and collaboration in the area of Computer Vision and Artificial Intelligence. The core of GeoGen3DHuman is on developing mathematically principled generative frameworks for deep learning on non-Euclidean domains such as graphs and 3D meshes. GeoGen3DHuman touches upon some of the most challenging problems in different fields such as computer vision and graphics, where generative models are very crucial. The research topic itself is very timely in terms of need and applicability of the systems targeted. This research also seeks to advance fundamental tools, that are not only of high relevance in terms of intellectual merit but also in broad impact.

Specifically, it develops techniques for geometric deep learning on 3D meshes, generative models in non-Euclidean domains and applications that use 3D models of the face and of the human body

Keywords: artificial intelligence, geometric deep learning, 3D/4D human.

The IRP Machine Learning, Network, System and Security (MLNS2) is interested in cybersecurity, which is a crucial research topic both in Cameroon and in France. It's mainly interested in two problems: the proliferation of malware on smartphones and phone call fraud that several African countries suffer from.

Created in 2022, MLNS2 associates in France CNRS and several laboratories namely Laboratoire d'Informatique en Images et Systèmes d'Information (LIRIS - CNRS/INSA de Lyon/Université Claude Bernard Lyon 1), Laboratoire d’Informatique de Grenoble (LIG, CNRS/Université Grenoble Alpes), Institut de recherche en informatique et systèmes aléatoires (IRISA - CNRS/Université de Rennes 1), and in Cameroon the University of Yaoundé I and its computer science laboratories.

Keywords: security, operating system, machine learning, networks, privacy.

The Montpellier-Stanford Robotic Research Project (MS-2RP) is a partnership between the Montpellier Laboratory of Computer Science, Robotics and Microelectronics (LIRMM - CNRS/Université de Montpellier) and Stanford University. Its objective for the coming years is based on collaborations based on the three main axes presented below: 
• underwater robotics, 
• medical robotics, 
• semantic web. 
Ultimately, this collaboration aims to expand to other topics such as human-robot interactions, biomedical applications, data science, etc.

Neuromorphic engineering aims to design computing architectures inspired by the structure and functioning of biological neural systems, based on spiking neurons and event-driven information processing. Their massive parallelism, low energy consumption and remarkable adaptive capabilities could revolutionise the way we control dynamic systems. In this context, the objective of the IRP NEUROCON is to develop methodological tools derived from the theory of hybrid dynamic systems for neuromorphic control. The project began in January 2026 and brings together members of the Centre de recherche en automatique de Nancy (CRAN - CNRS/Université de Lorraine) and Eindhoven University of Technology (Netherlands).

Project leaders: Romain Postoyan, CNRS Research Director at CRAN and Maurice Heemels, Professor at Eindhoven University of Technology.

Key participants: J. Daafouz, M. Giaccagli, M. Jungers, J. Kreiss, J. Lohéac, P. Lorenzetti, I.C. Morarescu (CRAN); E. Petri, F. Shakib, E. Steur, S. van den Eijnden, N. van de Wouw (TU/e).

The PhraseoPrag IRP (Modelling Everyday Conversation’s French and Japanese Pragmatic Phraseologisms) (2026–2031) is based on an equal, international and multidisciplinary collaboration (French and Japanese linguistics, NLP, educational AI) between five teams from seven French and Japanese universities: the Laboratoire d'informatique de Paris-Nord (LIPN - CNRS/Université de Sorbonne Paris Nord), LIDILEM, the University of Grenoble Alpes, the University of Lyon Jean Moulin, the Institute of Global Studies, Tokyo University of Foreign Studies, the University of Fukuoka, the Centre for Education (Faculty of Innovative Management Science), Chiba Institute of Technology, the Research Faculty of Media and Communication (RFMC), and the University of Hokkaido.

The project aims to develop a multilingual (French–Japanese) database dedicated to idioms, designed to serve both as a lexicographical resource and as a corpus for use in natural language processing (NLP). The resource, accessible to a range of users (learners, teachers, translators and researchers), will offer multimodal descriptions based on authentic data (spoken interactions, social media exchanges, etc.), structured and annotated for computational use. Finally, the project explicitly targets NLP applications: the collected and enriched data will enable the training, adaptation and evaluation of models for complex linguistic phenomena, notably idiomatic expressions and non-standard varieties, particularly in contexts of everyday interactions, whether spoken or mediated.

Project leader: Aude GREZKA, research engineer at LIPN

Artificial intelligence (AI) in healthcare promises more precise and personalised medicine, but its clinical adoption remains limited. The main barriers are a lack of generalisability, low robustness and the opacity of decision-making. The IRP on the Reliability and Transparency of Artificial Intelligence Models in Healthcare (SANTAI) aims to develop innovative AI solutions that incorporate knowledge specific to the medical field. The objective is to enhance the generalisability and transparency of decision-making mechanisms, which are considered key drivers for ensuring reliable clinical translation. The methods developed will be evaluated in the context of diagnosing cardiac and neurodegenerative conditions, two areas with a significant impact on public health.

Established in 2026, SANTAI brings together, in France, the Centre de recherche en acquisition et traitement de l'image pour la santé (CREATIS-CNRS/INSA Lyon/Inserm/Université Claude Bernard Lyon 1) and, in Canada, the VITALab and SNAIL laboratories, affiliated with the University of Sherbrooke.

Project leader: Olivier Bernard, professor at INSA Lyon and member of CREATIS

The research carried out within the framework of the IRP Systems, Verification, Fundamental Training, LogIque, Statistics (SINFIN) focuses on the use of formal methods in the implementation of theories and automatic tools for modelling, verification and development of complex software.

Created in 2019, SINFIN succeeds the LIA Infinis which started in 2011. It associates the CNRS, the Université Paris Diderot, the Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) and the University of Buenos Aires.

Keywords: fundamental computing, Logic, Languages, Verification and Systems

Warsaw, Paris, and Bordeaux are major research centers in automata theory, logic, and game theory. The InternIRP) Le Trójkąt aims to structure and develop these collaborations, strengthening the historical ties between France and Poland in these fields. Since the 1990s, figures like Damian Niwinski, Igor Walukiewicz, and André Arnold have significantly contributed to these exchanges. Recent successes include solving complex problems such as the complexity of reachability in Petri nets and parity games, thanks to the work of researchers from the University of Warsaw, the l'Institut de recherche en informatique fondamentale (IRIF - CNRS/University of Paris) in Paris, and the Laboratoire bordelais de recherche en informatique (LaBRI - CNRS/Bordeaux INP/University of Bordeaux) in Bordeaux. The IRP Le Trójkąt seeks to expand these collaborations beyond the Paris-Bordeaux-Warsaw triangle, involving other institutions in France and Poland. Through organizing scientific events and supporting research, the project fosters cooperation to tackle major challenges in fundamental computer science.