5th International Conference on Spoken Language Processing
We have built a multimodal-input multimedia-output guidance system called MMGS. The input of a user can be a combination of speech and hand-written gestures. The system, on the other hand, outputs a response that combines speech, three-dimensional graphics, and/or other information. This system can interact cooperatively with the user by resolving ellipses/anaphora and various ambiguities such as those caused by speech recognition errors. It is currently implemented on a SGI workstation and achieves nearly real-time processing.
Bibliographic reference. Takezawa, Toshiyuki / Morimoto, Tsuyoshi (1998): "A multimodal-input multimedia-output guidance system: MMGS", In ICSLP-1998, paper 0958.