Our Mission
The Department of Intelligent Multimedia Interaction (IMI) at IBM T. J. Watson is a research group, concentrating on building next-generation intelligent information systems that exploit multimodal, multimedia human-computer interaction. In particularly, we use multimodal to refer to various input modalities that users can use to express themselves to a computer, including natural language, direct manipulation, deictic gesture, and GUI. On the other hand, we use multimedia to refer to all output channels that a computer can use to respond to users' input, including animated 2D/3D visual presentations, speech/text, and video.
Our work lies in the heart of an interdisciplinary area known as Intelligent User Interaction (http://www.iuiconf.org). In particular, our team’s effort is centered around the development of new paradigms, methodologies, and metaphors to enable intelligent human-computer interaction, currently in the context of assisting users in their information access and analytic tasks. Using a combination of input modalities, including natural language and GUI, users can express their information requests naturally and precisely. Based on a user’s information request, computers on the other hand can interpret user requests in context and automatically generate customized multimedia tour of information to help users comprehend rich information easily. Ultimately, our work enables a new interaction paradigm, where users and computers are engaged in a continuous, progressively evolving multimedia discourse, during which the users and the computers work together cooperatively and learn from each other to leverage the strengths of both humans (e.g., abstract reasoning) and computers (e.g., sheer computing power) and help humans to accomplish their tasks efficiently and effectively.
Our Strength
Due to the interdisciplinary nature of our mission, IMI is made up of researchers from multiple research disciplines. In particular, our strength lies in the following areas: intelligent user interfaces, information retrieval, multimodal dialogue systems, automated multimedia authoring, natural language processing and generation, pervasive UIs, and 3D graphics user interfaces. In addition to the close intra-departmental collaboration, we have also established collaboration ties with researchers within and outside of IBM research.
Last updated 28 Dec 2007
