Looking for Research Scientists and Research Engineers for research topics in AI for Healthcare



  • Looking for Research Scientists and Research Engineers for research topics in AI for Healthcare


  • We have open positions for research scientists and research engineers at the University of Strasbourg & IHU Strasbourg Hospital.


  • - We are looking for research scientists to conduct novel research in AI for healthcare in the areas of medical image analysis and computer aided surgery. The successful candidates will contribute to developing new areas of research at IHU Strasbourg and mentor a growing team of PhD students and engineers. Salary is competitive and permanent contracts will be offered to strong candidates with experience.


  • - We are looking for experienced research engineers in the areas of medical image analysis, computer aided surgery and federated learning to support our research team.


  • - We also have multiple internship positions in computer vision and machine learning for healthcare.


  • The successful candidates will be hosted within the AI team at IHU Strasbourg/University of Strasbourg, an institute offering an international environment with state-of-the-art computing resources and unique clinical facilities for both patients and medical research. More information about the positions is available here.


  • Nicolas Padoy Professor of Computer Science, University of Strasbourg Director of Computer Science and AI Research, IHU Strasbourg Head of Research Group CAMMA University of Strasbourg / ICube IHU Strasbourg, 1 place de l'Hôpital 67000 Strasbourg, France


  • Web: http://camma.u-strasbg.fr


  • Phone: +33 (0) 3 904 13530


Postdoc offer: Prediction of driver loss score from data telematics



  • Duration: 1 year


  • Start: As soon as possible


  • Salary: €55,000


  • Institution: Polytechnic Institute of Paris / Telecom SudParis Plaiseau, France Supervisor: Prof. Mounim A. El Yacoubi


  • Professor, HDR Polytechnic Institute of Paris / Telecom SudParis / Mines-Telecom Institute SAMOVAR, CNRS, UMR 5157


  • Address: 19 place Marguerite Perey 91120 Palaiseau, France Phone: (+33) 1 75 31 44 53 (+33) 6 83 70 12 89


  • email: [email protected]


  • web: https://elyacoubi.wp.imt.fr/


  • Candidate:


  • - Master 2 or engineer in Data Science, Artificial Intelligence, Signal and Image Processing, etc


  • - Solid experience in Machine Learning, in particular Deep Learning, supervised techniques and unsupervised


  • - Excellent analytical skills for solving real problems


  • - Excellent programming skills (Python)


  • - High level publications


EPITA is opening several full-time computer science teacher-researcher positions



  • EPITA is opening several full-time computer science teacher-researcher positions, for recruitment no later than the beginning of the 2022-2023 school year.


  • In order to support the School's development dynamics at the national level, positions are to be filled on the sites of:


  • - Paris (Kremlin-Bicêtre and Campus Cyber à la Défense), - Lyon, - Rennes, - Strasbourg, - Toulouse,


  • to consolidate our teams and research axes on the following themes:


  • - Security of software and architectures: identification, protection, detection and reaction,


  • - Low-level system (kernel, assembler), operating systems, virtual machines and cloud computing,


  • - Embedded system (including robotics),


  • - Data science and engineering, knowledge extraction,


  • - Machine learning and other subfields of AI,


  • - Image processing, pattern recognition and vision,


  • - Automata and their applications (including verification and synthesis),


  • - Software and performance (including HPC, GPU).


  • It is not formally necessary to have the qualification for the positions of lecturer or university professor to be able to apply.


  • The precise information concerning these positions and the link to transfer your application to us are available here:


  • - https://www.lrde.epita.fr/~theo/postes_EPITA_MCF_2022.pdf for MCF profiles,


  • - https://www.lrde.epita.fr/~theo/postes_EPITA_HDR_2022.pdf for HDR profiles or very soon HDR.


  • The deadline for applications is April 15, 2022.


  • (The recruitment procedure can be read here: https://tinyurl.com/ProcedureRecrutementEPITA2022)


  • Sincerely, Pierre Parrend, EPITA Safety and Systems Laboratory (LSE) / ICube Unistra


  • Assia Touil, Director of Studies,


  • Thierry Géraud, Director of Research


  • http://www.epita.fr/


  • EPITA is a private school of computer engineers with CTI accreditation since 2007, evaluated by Hcéres (the last wave was in 2017-2018, the next one is in 2024-2025), and attached to the Doctoral School "EDITE de Paris" (ED 130).


Post-doc in multi-actor modeling



  • Participatory co-design of SMA models and simulations for multi-stakeholder decision support


  • Duration: 18 months


  • context


  • As part of the “Action network 2: Restoring values ​​to the grassland”, the platform multi-actor FAO “Global Agenda for Sustainable Livestock” (FAO-GASL-AN2), an activity conceptual modeling conducted by researchers from many countries (Argentina, Brazil, Canada, France, Senegal, Vietnam, Mongolia, New Zealand) led to the identification of 4 dimensions (productive, social, environmental and (development local) offering a framework for analyzing the multi-functionality of livestock farming.


  • This conceptual modeling resulted, among other things, in a first model (LivestockScape) to simulate, manipulate and visualize these different dimensions on abstract territorial configurations. This model integrates simplified soil processes (carbon, nitrogen), biomass dynamics (crops, meadows), herd productivity (zootechnics) and their management, strategies individual and collective land use planning and production organization, as well as production infrastructure and markets.


  • It is therefore the result of very strongly interdisciplinary work. Different lands and projects have expressed their wish to specialize this model to their context in order to be able to animate multi-actor discussion platforms (producers, decision-makers, NGOs, etc.). The realities socio-spatial aspects of countries such as Senegal, France (PACA region) and/or Vietnam, need to integrate other activities into the model (forestry, fodder crops or other, etc.), and other issues.


  • Among these we find, for example, the question of the carbon sequestration in Senegal as part of the DSCATT project. Furthermore, a initiative called DeMoCo (Collaborative Modeling Approach) funded in 2020 and 2021 aims to standardize the documentation of the models allowing the co-design of these ci and the follow-up of their evolution according to the questions addressed.


  • Subject of the post-doc The objective of this project is to design and test a participatory device, made up of methods and tools to surround the process of co-construction of models as well as co-exploration of models (scripting) by a multiplicity of actors heterogeneous.


  • The starting point is the identification of heterogeneous actors in terms of their scale of intervention and their objectives, having to co-exist on the same territory. Upstream, one of challenges is to allow these heterogeneous actors to participate in the design of the model according to their own points of view, to better understand the functioning of the territory that they share.


  • Downstream, the challenge is to provide the means for these actors to explore a variety of scenarios (possible futures) and options (possible decisions), according to their different points of view, in order to collectively assess the impacts.


  • The methods and tools developed will be tested within governance platforms of territories in the North and in the South and will serve as a support for discussions between the actors to equip anticipation processes. These issues and the associated objectives raise questions of comprehensibility and usability of the model to animate the debates outside the inner circle of specialists.


  • It's about therefore to design in a participatory way a set of sub-models (design) and dashboards (exploration), like so many points of view on a shared territory and thus allowing the expression by each of the stakeholders of desirable states; the carbon sequestration, the sustainability of the territory, household income, being considered like so many levers that can be activated. Technically, it will therefore be a question of considering the model in an agnostic way (black box, without a priori on the platform).


  • Upstream, these different points of view will be integrated into the functioning of the model, and downstream different dashboards will be connected to this same model in order to be able to show the interactions between the decision spaces of the different actors. This work will contribute to the reflection on the documentation of models (DeMoCo), and its use to derive explicit models and interfaces, which can be appropriated by multiple non-specialist actors. This work also raises ergonomic problems and will be carried out iteratively, in constant interaction with the actors of the different fields of experimentation.


  • Investigation protocols will be developed to assess the ease with which the different actors will be able to appropriate the functioning of the model, and understand the results that are produced. The results of these surveys will be involved in the continuous process of improving the device.


  • Planned activities Organization of multi-stakeholder workshops, mainly in Senegal andt in France (PACA region). Development of a methodological canvas for the co-design and co-exploration of models. Development of a meta-model for the description of multi-point simulation models of view for generating the different simulation modules to be integrated into the model, and to generate the various dashboards for each of the actors identified as user of the model at least in the form of skeletons to complete. Scientific publications


  • Terms


  • We are offering an 18-month post-doc starting in 2022, based CIRAD in Montpellier within UMR Selmet and UMR SENS. The candidate will have a computer scientist profile, or modeler (agronomy, geography, etc.) and skills in modeling and simulation multi-agent with a predilection/experience for participatory approaches to design.


  • The net salary will be between 2300 and 2900 euros net per month (depending on experience). The post-doc will require several trips to France and Senegal.


  • Required Skills


  • PhD in computer science, agronomy or geography with a strong content of modelization


  • Mastery of programming and possibly software engineering techniques Knowledge in modeling and in particular in multi-agent systems


  • Openness to participatory co-design approaches


  • Interdisciplinary experience


  • Required qualities


  • Ability to listen, ability to abstract, autonomy, teamwork. Mastering French and English, both orally and in writing.


  • To inquire [email protected] or [email protected]


  • To apply


  • You must send i) a detailed CV, and ii) a cover letter to jean- [email protected] before April 30, 2022.



Dynamic behavior evaluation by artificial intelligence Application to the analysis of the safety of the dynamic behavior of a vehicle



  • 1 Description of the thesis topic


  • The main goal of this thesis project is to develop a design methodology for a system of hybrid Artificial Intelligence mixing model-based approaches with those based on data for evaluating the dynamic behavior (comfort, safety, performance) of mobile systems inhabited. The application of interest to the DGA is the analysis of the safety of the dynamic behavior of military vehicles.


  • It is part of the field of study related to vehicle dynamics and its interpretation in particular applications. In manned mobile vehicles, the analysis and characterization of comfort, performance analysis, safety, characterization and monitoring of driving modes, represent many different subjects of study, but which have common characteristics. They do involve dynamic interactions with the environment, a problem that is difficult to model.


  • Thus, the center of concern of the study relates to directly measurable characteristics or indirectly estimable, today calling on particular human expertise. Another one main characteristic is the amount of data available which is not excessive. The evaluation of a manned mobile system is a complex problem because it mixes both sciences (vehicle/environment interaction, vehicle dynamics, seat dynamics, and science human (on feelings, sharing control, awareness of situations).


  • It is clear that for a such complexity of transmitting dynamic phenomena, model-based approaches reach their limits in the successive approximations made for the establishment of the models. In complement, data-based approaches add robustness when the data used during learning are sufficiently representative of the targeted area of ​​use. The The methodology proposed will be hybrid and should make it possible to use the best of both approaches.


  • Desired profile of the doctoral student Knowledge of algorithms, artificial intelligence, robotics, Team work, Taste for field experimentation.


  • The thesis being co-financed by the Ministry of the Armed Forces within the framework of the Call for projects (AAP) 2022 classic AID theses from the Defense Innovation Agency, the doctoral student must be:


  • • Citizens of the European Union, the United Kingdom or Switzerland


  • • Holders of a master's degree or equivalent, or an engineering degree conferring the master's degree


  • • Available 1 October at the earliest and 1 December at the latest


  • Supervisory team


  • INRIA thesis supervision: Philippe Martinet, Research Director ACENTAURI team, Inria Sophia-Antipolis [email protected]


  • Co-supervision: Nicolas GUTOWSKI University of Angers LERIA, EA 2645 [email protected]


  • Contact DGA: Sebastien AUBIN AI and BigData expert DGA TT/SDT/MMP/MIA [email protected]


  • Organization


  • This section presents the organization of this project by detailing the steps planned for each year. The thesis being a collaboration between different laboratories, the doctoral student will be bi-located between INRIA in Sophia-Antipolis and the DGA site of Angers, on INRIA initially for about 1.5 years then in Angers for the remainder of its work.


  • Application


  • The candidate must send a CV, a motivation letter, his report card of the current year and the previous year, before 03/25/2022, 2 letters of recommendation, including at least the opinion dated and signed by the director of his last training, putting in the subject of his email “[INRIA/LERIA/DGA] Application” to the three supervisors.


post-doc offer in web mining University Grenoble Alpes



  • Skills - knowledge in software engineering and advanced experiences in development and architecture python software, - knowledge of AI technologies (clustering, classification, etc.),


  • - desired knowledge in digital social network analysis, - Desired knowledge of natural language processing (NLP) and/or statistics lexical,


  • - knowledge and practice of server administration, - knowledge of NoSQL databases, - Know Gitlab or Github or even Docker,


  • - desired skills in web development and interfaces (PLOTLY, FLASK, etc.), - team work.


  • In the context of the COPERNIC prematurity project (CNRS - Grenoble-Alpes University) you will contribute to the development of a framework for the supervised characterization of flow publications from the Web and social media, in an information monitoring context.


  • This framework embeds social network analysis and textual analysis technologies (TAL, Text mining). In collaboration with the members of the project, you will participate in the evolution of the architecture functional (3 tiers) and qualitative of the framework from an industrial and scientific perspective open (FAIR).


  • You will intervene in representational engineering (Mongodb, redis,...) and software in order to ensure the qualities of robustness and interoperability of the framework. The developments are carried out in a python environment on a linux server (ubuntu).


  • Activities - administration of the linux development / production platform,


  • - develop strategies for testing and evaluating developments in the context of experiments (use boxes),


  • - ensure the update and curation of shared development versions (GitLab),


  • - contribute to the implementation and respect of the principles of science and open data,


  • - develop the software packaging and the technical documentation of the framework,


  • - python development of software modules and associated APIs,


  • - contribute to the implementation of demonstrators, - participate in the animation and the life of the team.


  • Job offer Post-Doc Big data development, Web mining M/F


  • Work context The candidate will work within the Research Federation "Innovation, Knowledge and Society (FR3391 Innovacs) in conjunction with the GRICAD team at the Maison de la création et de l'innovation (MaCI). Under the supervision of UGA, Grenoble - INP and CNRS.


  • The FR Innovacs brings together 18 laboratories in human, social and engineering sciences. It helps these laboratories to develop research multidisciplinary around innovation issues. It aims in particular to study the impact of innovations on society.


  • Website: https://innovacs.univ-grenoble-alpes.fr


  • GRICAD (Grenoble Alpes Research Intensive Computing and Data Infrastructure) is a Support Unit in research (UAR3758) under the supervision of the CNRS (INSMI), the University of Grenoble Alpes, Grenoble - INP and INRIA, created in 2016, to meet the needs relating to the calculation and data of researchers from the Grenoble site. Website: https://gricad.univ-grenoble-alpes.fr


  • Constraints and risks: Sedentary work on screen.


  • Please submit your CV and cover letter to post to [email protected]


  • General informations Duration of the contract: 15 months - Work shift: Full time


  • Remuneration: 2663 and 3069 € gross


  • Desired level of studies: Doctorate - Desired experience: 1 to 4 years



Sorbonne University: 4 ATER positions in computer science (closing of applications on 5 April)



  • The Faculty of Letters of Sorbonne University offers several POSITIONS of ATER (Artificial Intelligence profile for the Human Sciences) within the UFR of Sociology and Computer Science for the Human Sciences. 9 permanent staff and 5 ATER teach in the computer science department of the UFR.


  • The courses take place in 2 sites of the Faculty of Arts in Paris (Clignancourt Center, Malesherbes Center) as well as at the Maison de la Recherche (28 rue Serpente, St Michel district) where the UFR and the ATER office are located.


  • The list of teachers of the department appears here http://lettres.sorbonne-universite.fr/faculte-des-letters/ufr/human-and-social-sciences/sociology-and-computer science-for-1), we are mostly attached to the Computational Linguistics team of the STIH laboratory (list of members and list of recent publications).


  • The courses are mainly in the Computer Science part of the Bachelor's Degree in Language Sciences and in the Master's degree in Language and Computer Science. Among the needs are programming (Python, C++, Web), Python for NLP, corpus linguistics, Computer Networks and Databases. Part of the service will concern digital training "PIX" (http://www.pix.fr/)


  • In terms of research, if the profile is not closed, we are looking primarily for the themes of the "Computational Linguistics" team, namely: Automatic Language Processing/Computer Linguistics, Speech Processing, Knowledge Engineering. The team's recent projects (ANR in particular) could also lead to interest in image processing and AI profiles in the broad sense.


  • The procedure is open from Tuesday 15 March 10 am and will close on Tuesday 5 April at 4 pm.


  • The position will be on September 1, 2022


  • The detailed application conditions can be found here:


  • https://recrutement.sorbonne-universite.fr/fr/personnels-teachers-researchers-teachers-researchers/teacher-researchers/campaign-of-recruitment-ater. HTML


  • The list of positions is available here: https://recrutement.sorbonne-universite.fr/fr/personnels-teachers-researchers-teacher-researchers/teacher-researchers/recruitment-campaign-ater/positions-opened-by-the-faculty-of-letters.html


  • NB: apply please on the SUFLATER47 profile (which actually has 3 positions) and on the SUFLATER48 profile to facilitate the ranking


  • Teaching contacts: - License: [email protected], [email protected]


  • - Master: [email protected], [email protected]


  • - Pix: [email protected] Research contacts: [email protected]


  • Administrative contact: [email protected]


Recruitment of 2 ATER in Computer Science - IUT de Laval (Le Mans University)



  • The IUT of Laval, component of Le Mans University offers two ATER positions in computer science.


  • The courses take place either in the COMPUTER department or in the MMI department (Multimedia and Internet Professions), on the site of the IUT of Laval.


  • The precise job descriptions are on Galaxie: - Computer Science Department: https://www.galaxie. teachingup-research. gouv.fr/ensup/ ATERListesOffresPubliees/0530951W/FOPC_43947.pdf


  • MMI Department: https://www.galaxie. teachingup-research. gouv.fr/ensup/ ATERListesOffresPubliees/0530951W/FOPC_43948.pdf


  • In terms of research, the profile is not closed, we are looking for people who can integrate into the activities of the IEIAH team at LIUM.


  • The procedure is open from Tuesday 15 March 10 am and will end on Tuesday 5 April at 4 pm.


  • The start of the post will be on 1 September 2022


  • Teaching contacts: - BUT IT: [email protected] - BUT MMI: [email protected]


  • Contact research: [email protected]


  • Administrative contact: applications-ater@univ-lemans. fr


PhD Position Computer Sciences/ Tours University, France. Integration of combinatorial post processing in learning architectures



  • Vacancy: PhD position PhD Thesis subject: integration of ad-hoc combinatorial post processing in learning architectures


  • Supervised by: Hugo Raguet, Julien Mille and Romain Raveaux ([email protected], [email protected] and [email protected]) from the LIFAT laboratory of Tours in France (https://lifat.univ-tours.fr/).


  • Description: Many prediction tasks exhibit some kind of spatial regularity, e.g. semantic segmentation of images or 3D point clouds since most neighboring pixels or points belong to the same class. In cases for which state-of-the-art methods involve deep neural networks, it is common to resort to a postprocessing regularization step to enhance the quality of predictions. In contrast, we would like to investigate the integration of such regularization step within the neural network architecture.


  • This is motivated by two potential benefits. First, this would make the neural network aware of the regularization process during the learning phase, possibly leading to better prediction but more importantly to learn from smaller or corrupted training data. Second, this would enable the learning of the regularization parameters themselves, making the regularization step more adaptive.


  • This approach can be generalized to other kind of regularization or post processing, which are currently used in situations where experts must introduce prior knowledge for learning adequate representations, typically due to lack of data. More often than not, such regularization are combinatorial by nature.


  • The main difficulty is to find regularization methods with acceptable computational load and that can be differentiated with meaningful derivatives. Sometimes, the combinatorial nature can be relaxed to continuous formulations, for instance ℓ1 regularization for feature selections or total-variation regularization for spatial regularity, leading to a process that is essentially differentiable. Recently, new techniques [1] introduced ways of differentiating through the actual combinatorial solver. Our goal is to explore and compare both techniques.


  • The starting point of our investigation would be graph-based spatial total variation regularizations, solved using our parallel cut-pursuit approach [2]. This would the opportunity for close collaboration with Loic Landrieu at IGN (Paris), being a still pending task at the heart of his project READY3D [3].


  • The PhD will be paid with legal allowances (about 1600 euros).


  • Please submit pdf files of your CVs along with any meaningful information for the position (motivation letters, reference letters, code source, academic records, awards, articles, ...)


  • If your application is selected, you will then be contacted for further information and interview details.


PhD offer: MACHINE LEARNING APPROACHES APPLIED TO MONITORING OF ENERGY CONSUMPTION AND ENVIRONMENTAL DATA



  • THESIS DIRECTOR: PATRICE WIRA CO-SUPERVISOR: GILLES HERMANN COMPUTER RESEARCH INSTITUTE, MATHEMATICS, AUTOMATICS AND SIGNAL (IRIMAS), IMTI TEAM, 61 RUE ALBERT CAMUS, 68093 MULHOUSE CEDEX TEL: 03 89 33 74 11 / E-MAIL: [email protected]


  • Today, sensors can be installed everywhere to measure the environment and the uses of a building. The measured data can be centralized, processed and analyzed by algorithms in order to react quickly accordingly or to predict. The use of mathematical tools, signal and data processing, and the Machine Learning (set of methods/models capable of learning to solve a problem from the data) make it possible to achieve this [1].


  • This research project aims to exploit electricity consumption and water consumption of one or more rooms or an entire building. These data can be supplemented by external environmental data (temperatures, pressure, brightness, wind, etc.) and internal (opening of doors, passages, air flow, etc.).


  • These signals and consumption directly reflect the activity of uses and are impacted by the environmental data [2]. These signals and data are of very different natures each other and are difficult to exploit together, in a quantitative way [3].


  • The objective of this work is to develop tools for visualization, analysis and understanding of these different heterogeneous quantities measured in an environment closed and occupied by users.


  • The massive character (especially in terms of dimensionality) and the diversity of the quantities measured leads to a form of complexity in their analysis.


  • Primitives representative of the different groups of uses or users will be used in supervised learning approaches to recognize said uses or users. The behaviors can be characterized by measurement.


  • Of all the indicators, measured or calculated, available only the most relevant should be retained, with a view to real-time operation.


  • We will study the application of existing machine approaches learning [4], supervised and unsupervised, such as artificial neural networks, time series, probabilistic approaches to determine modes of operation or types of use.


  • We will also seek to explore their limits and compare them to more classical deterministic approaches to propose new methods.


Postdoc Offer on Temporal Answer Set Programming



  • Supervisor: Mart ́ın Di ́eguez University of Angers, Angers, France email: [email protected]


  • During the last years several hybrid extensions of Answer Set Programming (ASP) [1] have been published. The hybrid extensions we are interested in are temporal [2] and constraint Answer Set Programming [3].


  • The former approach allows reasoning on time scenarios using temporal operators from Linear Time Temporal Logic [4]. The latter approach combines ASP with Constraint Programming (CP) approaches in order to improve the performance of the ASP solvers.


  • In this project we want to consider the integration of constraints in the context of temporal ASP. In the classical case, definitions of linear time temporal logics built on the top of a constraint solving system have been proposed in the literature [5]. Although those logics are undecidable, some fragments meeting some completion condition [6] turned to achieve decidability.


  • We are interested in combining constraints systems and temporal ASP by using Equilibrium Logic [7], a logi- cal characterisation of ASP, which led to the most prominent extensions of ASP with temporal and constraint concepts [2, 8].


  • The duration of the contract will be of 18 months, the starting date is expected to be in June and the salary will be around 2158,79 e net.


  • The research lab LERIA1 (laboratoire d’ ́etude et de recherche en informatique d’Angers) is one of the research units at the University of Angers. It is located at the UFR Sciences campus in Angers (Pays de la Loire), France and it is run by Eric Monfroy2 .


  • The research activity of the lab relies on the three main axes:


  • • Metaheuristics and combinatorial optimisation (MOC, in French)


  • • Reasoning under uncertainty and constraints (RIC, in French)


  • • Artificial learning and knowledge representation (ARC, in French)


  • Regarding this context, this project is related to the second research topic. Angers is a middle-sized city in western France, about 300 km (190 mi) southwest of Paris. It is the prefecture of the Maine-et-Loire department, well-connected with Nantes and Paris by high-speed trains.


  • In 2022 classification of the ”villes et villages o`u il fait bon vivre en France”, Angers is placed in the first position.


  • 3 Profile We are looking for PhD with the following requirements: • Solid background in: – Non-classical logics such as G ̈odel/modal/intuitionistic logic.


  • – Answer Set Programming and/or constraint programming.


  • – SMT and/or SAT


  • 1https://leria.univ-angers.fr/ [email protected]


  • 4 Documentation Candidates who are interested in applying, please send the following documentation to martin.dieguezlodeiro@ univ-angers.fr


  • • Short CV (3 pages maximum);


  • • The list of publications of the candidate;


  • • One contact willing to support the candidate.