Middleware Services for Multimodal Interactions in Smart Environments

Smart and intelligent environments can proficiently benefit from a new generation of multimodal dialog systems, which enable users to interact in natural ways like voice, gesture, facial motions, and so on. However, in order to make this possible, it is necessary to integrate such systems and services, possibly in a transparent way. In the present paper, we propose (i) an architectural model for multimodal smart environments; (ii) an agent-based middleware infrastructure that provides basic components for the integration of multimodal interaction systems; and, (iii) a tool for rapid-prototyping, which automatically customizes some middleware components to implement multimodal interaction mechanisms for the environment´s application services.