Diva, une architecture pour le support des agents gestuels interactifs sur internet

In this paper, we present a Web-based framework for interactive deictic and sign language virtual agents. Our framework is a DOM-Integrated Virtual Agents architecture, which has been first designed as a Web-based toolkit dedicated to the support of Embodied Conversational Agents. In this work, the toolkit is used both in the context of signed and vocal languages. We discuss the technical issues raised by the implementation of proper interactive Virtual Signers within our full DOM-Integrated approach, and we develop two studies: first, on coarticulation in dynamic sign language utterances generation; and second, on the way deictics are performed in both vocal and signed language. MOTS-CL

• Type: Journal
• Year: 2010
• Where: TSI
• Authors: Annelies Braffort, Jean-Paul Sansonnet, Jean-Claude Martin, Cyril Verrecchia
• Keywords: Artificial Intelligence, Language Virtual Agents, Sign Language, TSI 2010, Virtual Agents