Research Summary

My research focuses on the analysis, modeling, and recognition of social and emotional signals in the context of human-human and human-machine interactions (computer, robot, virtual agent, etc...). In the context of Robotics or Virtual Agents, my work concerns the perception part which designates the ability of the system to collect, process and format information useful to the robot/agent to act and react in the world around him.

My work is mainly based on computer vision, multi-modal fusion and automatic learning applied to the analysis of human behavior and the prediction of its state Emotional and/or mental by a machine.

In summary, I work on the automatic prediction of human characteristics (eg personality), the detection of human behavior (posture, gaze, etc. \dots), detection of the mental and emotional state of the human (Eg, detection of engagement, detection of facial actions and continuous prediction of emotion) in the context of a Man-Virtual Agent or Man-Robot interaction.

Interests

  • Artificial Intelligence
  • Robotics
  • Data Science
  • Affective Computing
  • Social Signal Processing
  • Human-Machine Interaction
  • Computer Science
  • Social Signal Processing
  • Image and Signal Processing
  • User-Centered design

Research Projects

  • image

    IMMEMO (IMMersion 3D basée sur l'interaction ÉMOtionnelle)

    In collaboration with an industrial partner who develops a 3D immersion system consisting of a network of synchronized cameras, display screens, software to capture the movement of a person in a restricted environment and to synthesize in real time A clone of the person immersed in the virtual world. This industrial research project aims at designing an extension module enabling this platform to capture and recognize facial expressions of the immersed person (learner) in order to facilitate his interaction with an autonomous conversational agent whose behavior will adapt Semi-automatically to that of the learner.

    http://www.rennes.supelec.fr/immemo/

  • image

    Investissement d'Avenir A1:1 – Interaction avec les avatars à taille humaine

    The objective of this project is to allow an interaction between a human and a virtual agent of human size, in a cultural place (the Historial of the Vendée), including dimensions of verbal and non-verbal communication. The avatar plays the role of guide who asks visitors about the museums in order to acquire new knowledge. To interact properly with visitors to the museum, the virtual agent must have communication skills to interpret the verbal and non-verbal cues of the users.

  • image

    IRQN (Images Qualité du Réseau routier National)

    Each year, as part of Operation IQRN, Images Qualité of the National Road Network, one third of the national roads are evaluated in order to determine the quality of the pavement and plan the necessary repairs. The manner of detecting cracks was based on a manual survey by qualified persons (lifters). This task is dangerous for the lifters, tedious, little repeatable, ineffective. This project involves the automation of pavement degradation surveys.

Principal co-organized international events

  • 2018
    The General Data Protection Regulation: An opportunity for the CHI community? at CHI 2018, Montreal, Canada
  • 2015
    INTERPERSONAL "First International Workshop on Modeling INTERPERsonal SynchrONy" at the International Conference on Multi-modal Interaction (ICMI), Seattle, USA
  • 2015
    First international Workshop on Engagement in HumAN Computer IntEraction" at 6th Affective Computing and Intelligent Interaction (ACII), Xian, China