Conception et évaluation d'avatars hybrides

par Adélaïde Genay

Projet de thèse en Informatique

Sous la direction de Martin Hachet et de Anatole Lecuyer.

Thèses en préparation à Bordeaux , dans le cadre de École doctorale de mathématiques et informatique (Talence, Gironde) , en partenariat avec Laboratoire bordelais de recherche en informatique (laboratoire) et de Images et Son (equipe de recherche) depuis le 15-10-2019 .


  • Résumé

    Cette thèse est dédiée à la conception de nouveaux paradigmes d'interaction permettant aux utilisateurs de contrôler leur avatar de la manière la plus naturelle et efficace. L'objectif final est de renforcer le sentiment d'incarnation de l'avatar des utilisateurs en leur offrant un moyen d'interaction aussi transparents que possible. Pour atteindre cet objectif, nous explorerons des approches d'interaction à la fois directes et indirectes que nous évaluerons ensuite.

  • Titre traduit

    Design and evaluation of hybrid avatars


  • Résumé

    In this PhD, we will explore new forms of avatars that are physical, or a combination of real and virtual components, in addition to screen-based display only. Indeed, it has been shown that physical characters may enhance social presence compared to pure virtual avatars [Hornecker 11]. First attempts of such physical avatars have notably been experimented by Potioc in bio-feedback and education scenarios [Gervais 16]. We propose to explore in more depth such tangible avatars, as well as avatars dedicated to augmented reality setups (e.g. Hololens). In particular, we will first concentrate on the design, development and evaluation of hybrid avatars that allow one to perceive the current state of a remote collaborator (e.g. the shape and appearance of the tangible avatar show me that my collaborator is active, sleeping, stressed, and so on). Second, we will invesgate how such hybrid avatars can be used to modify one's own avatars (e.g. moving the physical avatar's hand to attract the collaborators attention, or moving the avatar away from the screen to indicate that I am not available). We will also explore AR-based techniques for interaction with avatars. To reach this general objective, we will start by exploring the state of the art. This will help us to identify the limitations of the current approaches. Then, we will build prototypes to test hypotheses. This implies software developments in the filed of AR VR, and interaction techniques. This will also imply developments that are linked to the underlying hardware technology (e.g. Hololens and tracking). Finally, we will evaluate our prototypes by conducting user studies.