Menu Close

ICAR Team: Image & Interaction


William PUECH
William PUECH


Image & Interaction


The ICAR (Image & Interaction) team brings together researchers from the two departments of Robotics and Computer Science around the theme of “image” and more generally visual data. It is currently made up of 9 permanent members, academics and CNRS, but also includes among its regular collaborators several university hospital doctors using medical imaging, researchers in remote sensing from the TETIS laboratory and in modelling for agronomy from CIRAD.

The ICAR team develops research themes combining the interaction and processing of visual data such as 2D, 3D, multi-spectral (nD) images, videos or nD+t image sequences and 3D objects, whether in the form of 3D meshes or parametric modelling.

The team is structured along 4 research axes

  • Analysis & processing;
  • Multimedia Security;
  • Modelling & Visualisation;
  • Artificial Intelligence for visual data.

Noura Faraj, Maître de conférences, UM
Emmanuel Faure, Chargé de recherche, CNRS
Marc Chaumont, Maître de conférences, U Nîmes
William Puech, Professeur des universités, UM
Frédéric Comby, Maître de conférences, UM
Christophe Fiorio, Professeur des universités, UM
Olivier Strauss, Professeur des universités, UM
Gérard Subsol, Chargé de recherche, CNRS

Associates & Students
Ange Clement, Société GEOMETRY FACTORY
Khelian Larvet, Stratégies
Erwan Reinders, Drone Geofencing
Kevin Planolles, UM
Tristan Cossin, TeraWT
Guillaume Fourret, Drone Geofencing
Mathieu Ladeuil, Game Source Studio
Robin Jarry, UM
Eugenio Dias Ribeiro Neto, UM
Marc Hartley, UM
Hugo Ruiz, UM
Thibault De Villèle, UM
Nicolas Dibot, CNRS
Cyril Barrelet, CNRS
Guillaume Picaud, CICAT Occitanie Montpellier
Martha Duraes, CHU Mtp

Regular Co-workers
Dorian Kauffmann, CDD Ingénieur-Technicien, CNRS
Nadine Jacquet, CDD Ingénieur-Technicien, CNRS
Tao Laurent, CDD Ingénieur-Technicien, CNRS
Benjamin Gallean, Invité longue durée Mission longue, CNRS
Guillaume Captier, Associé, UM
Magali Bourgoin, Doctorant externe, CHU Mtp
Guillaume Gay, CDD Ingénieur-Technicien, CNRS
Sylvain Leclerc, CDD Ingénieur-Technicien, UM
Yan Chastagnier, Doctorant externe, UM

The ICAR team is structured along 4 research axes:

  • The Analysis & Processing axis is interested in new low-level information processing techniques to improve the perceptible information in the image and to take into account, in the same theoretical framework, the imprecise, the uncertain and the incomplete (the different types of error in visual data processing).
  • The Multimedia Security axis is concerned with the securing of visual data. In order to ensure this security, coding algorithms are developed that combine watermarking, steganography, forensics, encryption and authentication and often require robustness to compression.
  • The objective of the Modelling & Visualisation axis is to model large sets of complex data (in dimension and nature) in order to allow intuitive visualisation or to manipulate these data to extract knowledge.
  • A transverse Artificial Intelligence axis for visual data has recently been set up. The team’s researchers have focused on promising Deep Learning methods and are developing new methodologies to design innovative network architectures or post-processing modules to exploit information from the last layers of a network, but also to implement coherent database construction procedures to optimise results.

Some of the team’s algorithms are available on the team’s website.

The ICAR team’s activity is divided between research, doctoral training (17 theses in progress) and industrial development, in particular via CIFRE agreements. In addition, all the members of the team are strongly involved in university training courses in the Occitanie-Est region (Béziers, Montpellier, Nîmes) related to images.

ICAR team awards and distinctions from 2017 to 2022 :

  • FUI RIDER project (2009-2013): the objective of the RIDER project (Reasearch for IT Driven EneRgy efficiency) is to develop an innovative information system to optimise the energy consumption of a building or a group of buildings. The consortium is made up of LIRMM, IES, PROMES (University of Perpignan), and industrial partners IBM, EDF, Cofely GDF-Suez, Pyrescom, Coronis, ASA and Enoleo.
  • NaturalPad (2013): NaturalPad, created by former LIRMM PhD students, develops serious games for therapeutic purposes, in collaboration with LIRMM and M2H, using consumer interfaces (Wii, Kinect, cameras).
  • ALCI (2012): created by former LIRMM PhD students, it specialises in robotics and vision. We enabled them to detect, only with camera control, multiple defects in the manufacture of clay roof tiles, such as cracks and creep. Thanks to these algorithms, ALCI was able to create a robot that is now on the production line.
  • CNRS PEPS SEMYO project (2012-2013): this project coordinated by the LIRMM, in partnership with the Pitié-Salpêtrière Institute of Myology in Paris and INRIA Grenoble, deals with the automatic analysis of MRI images of the human musculoskeletal system. The aim is to improve the diagnosis and computer-assisted monitoring of myopathies.
  • The VOODDO project (2007-2010) of the ANR Audiovisual and Multimedia: this allowed us to collaborate with the regional company Vodnet, the National Rugby League and the MAORE team of the LIRMM in order to protect the distribution of sports videos on the Internet.