Skip to Main content Skip to Navigation
Theses

Rendu multimodal en Réalité Virtuelle : Supervision des interactions au service de la tâche

Abstract : Our research deals with multimodal rendering in Virtual Reality (VR), in order to improve the interactions between users and virtual environments.
In particular, we try to make the most of human sensori-motor bandwidth and semantic capacities of visual, audio and haptic modalities, to increase comfort, performance or understanding of the operators in their different tasks. The ambition of our work lies in a generic and contextual approach of multimodal rendering in VR, in contrast with solutions tied to specific applications. Our research is also task-oriented rather than realism-oriented.
Our major contribution is the design of a multimodal distribution supervision process dedicated to VR applications. It is based on a generic model of interactions, which allows representing the semantics of the tasks and the rendering possibilities of the system, with the help of a limited number of simple elements. The core of the supervision process is a rule-based engine. It is in charge of deciding what is the most appropriate multimodal rendering for a given interaction, depending on a knowledge base and on the interaction context.
Evaluations are conducted on practical scenarios in the field of Fluid Mechanics exploration, Bio-Informatics protein docking and CAD. Results are satisfactory, in terms of the reliability of the decision process, its adaptability to various contexts as well as the speed of the answer, which is compatible with real-time VR interactions.
Document type :
Theses
Complete list of metadatas

Cited literature [115 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00193064
Contributor : Guillaume Bouyer <>
Submitted on : Friday, November 30, 2007 - 2:29:28 PM
Last modification on : Monday, December 14, 2020 - 9:46:45 AM
Long-term archiving on: : Monday, April 12, 2010 - 5:38:46 AM

Identifiers

  • HAL Id : tel-00193064, version 1

Collections

Citation

Guillaume Bouyer. Rendu multimodal en Réalité Virtuelle : Supervision des interactions au service de la tâche. Interface homme-machine [cs.HC]. Université Paris Sud - Paris XI, 2007. Français. ⟨tel-00193064⟩

Share

Metrics

Record views

517

Files downloads

890