Abstract

Computer vision-based sensing of people enables a new class of public multi-user computer interfaces. Computing resources in public spaces, such as automated, information-dispensing kiosks, represent a computing paradigm that differs from the conventional desktop environment and correspondingly, requires a user-interface metaphor quite unlike the traditional WIMP interface. This chapter describes a prototype public computer interface which employs color and stereo tracking to sense the users' activity and an animated speaking agent to attract attention and communicate through visual and audio modalities.

Introduction

An automated, information-dispensing Smart Kiosk, which is situated in a public space for use by a general clientele, poses a challenging human computer interface problem. A public kiosk interface must be able to actively initiate and terminate interactions with users and divide its resources among multiple customers in an equitable manner. This interaction scenario represents a significant departure from the standard WIMP (windows, icons, mouse, pointer) paradigm, but will become increasingly important as computing resources migrate off the desktop and into public spaces. We are exploring a social interface paradigm for a Smart Kiosk, in which computer vision techniques are used to sense people and a graphical speaking agent is used to output information and communicate cues such as focus of attention.

Human sensing techniques from computer vision can play a significant role in public user-interfaces for kiosk-like appliances. Using unobtrusive video cameras, they can provide a wealth of information about users, ranging from their three-dimensional location to their facial expressions and body language.