Avatar motion generation based on AI speaker response and timing

Abstract

This study proposes a system that works in conjunction with an IoT-based health management system using AI speakers to generate movements of avatars projected on walls and other surfaces, targeting general users and others unfamiliar with the operation of information devices. The health management system combines an AI speaker for voice control to manage daily personal activity data acquired from wearable devices. In this system, the avatar performs interactive actions by analyzing sentences spoken by the user when making inquiries and the AI speaker's responses. This paper describes the research background of this study, followed by an overview of the system concept proposed in this study, the development of a prototype system based on the concept, a preliminary test of the prototype developed in this study, and future issues for this study.