Interactive Embodied Agents for Cultural Heritage and Archaeological presentations

In this paper, Maxine, a powerful engine to develop applications with embodied animated agents is presented. The engine, based on the use of open source libraries, enables multimodal real-time interaction with the user: via text, voice, images and gestures. Maxine virtual agents can establish emotional communication with the user through their facial expressions, the modulation of the voice and expressing the answers of the agents according to the information gathered by the system: noise level in the room, observer’s position, emotional state of the observer, etc. Moreover, the user’s emotions are considered and captured through images. For the moment, Maxine virtual agents have been used as virtual presenters for Cultural Heritage and Archaeological shows.