Multimodal Deixis: Nearby Reality or Still Science Fiction?

A lot of research has already been done in the field of multimodal or natural interaction with computer systems. Present day interaction with a computer system must be learned on top of communication skills humans already possess. Studying new forms of interaction with computers might overcome this shortcoming, thus making human computer interaction more robust and fast. For this kind of interaction we need a computer that can see gestures, hear spoken words and make sense of them. This article will review the possibilities and problems regarding the different aspects of making computers see, hear and reason about multimodal deixis. We will see that though research in this field has come a long way, there are still some difficult obstacles on the road ahead