Just what are Perceptual User Interfaces, anyway? Who should be interested? And why this workshop? Well, to borrow from the workshop’s web site:

The general idea behind PUI is that current computer interfaces, which are almost exclusively GUI or WIMP-based, have significant limitations and make relatively little use of both human and machine perceptual capabilities. As computers become more ubiquitous and appear in a myriad of form factors and in all aspects of daily life, more natural and powerful interfaces will be a necessity. The goal of PUI is for people to be able to interact with computers in a similar fashion to how they interact with other people and with the physical world. To move in this direction, we need to involve several areas of study, such as speech and sound recognition, computer vision, haptic interfaces, language and dialog, learning, user modeling, human perception and cognitive studies. While the focus so far has been on input modalities, it is clear that output is just as integral to the effort. Not to mention what takes place "in the middle", e.g., reasoning and dialog management.

This is the second workshop devoted to the topic of Perceptual User Interfaces, or PUI. Last year the first such workshop was held in Banff, Canada in mid-October. In spite of the beautiful surroundings, tempting us to skip out and explore the wonderful Canadian Rockies in Banff National Park, we managed to corral about 67 people to stay inside and conduct a successful workshop. The meeting brought together a rather unique mix of participants, and generated significant enthusiasm. There appeared to be general agreement that we should do it again – this year we expect about 70 attendees to gather in beautiful San Francisco for the second "PUI Workshop."

In response to suggestions after PUI'97, we have tried to accommodate two particular requests in this year’s meeting: (1) a less busy schedule with more time for discussion, and (2) representation from a broader range of disciplines. However we still need to work toward a more diverse representation in the future.

While last year we published extended abstracts, this year we asked for full papers and research notes. These proceedings present the accepted workshop papers: nine full papers (for oral presentation), eleven research notes (presented in a poster/demo session), and five proceedings-only papers.

I would like to thank Microsoft Research for providing support and resources and Cindy Muszynski of American Express Travel, who was an enormous help in arranging for the workshop facilities and accommodations and serving as the focal point for online registration. Finally, many thanks to the program committee members, invited speakers, panelists, and session chairs for all their work.

All of the information in these proceedings, and more, is available at the workshop's web site: http://research.microsoft.com/PUIWorkshop.

Matthew Turk, PUI’98 Workshop Chair
November 1998