Developing a novel, scientifically well‐founded multimodal fusion and fission engine for smart spaces. This multimodal fusion and fission represents a core part of the architecture for future human-environment interaction systems and is domain‐independent. In comparison to existing approaches, the following extensions will be necessary: a. providing larger degrees of freedom for multiparty interaction towards a notion of multihuman‐ multi‐machine interaction; b. seamlessly integrating a multiplicity of devices and communication standards relevant for the smart spaces scenario; c. taking into account the fact that multiple agents (cars, respectively their drivers, passengers) interact with the same environment; d. supporting not only multiple displays inside the car but also time‐synchronized presentations in the outside environment; e. introducing extended cross‐modal reference (e.g. from inside to outside of the vehicle).
Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence