1 ) Affective Computing Maja Panitc Delft University of Technology, The Netherlands Copyright © 2005, Idea Group Inc., distributing in print or electronic forms without written permission of IGI is prohibited. INTRODUCTION We seem to be entering an era of enhanced digital connectivity. Computers and the Internet have be- come so embedded in the daily fabric of people’s lives that they simply cannot live without them (Hoffman et al, 2004). We use this technology to work, to communicate, to shop, to seek out new information, and to entertain ourselves. With this ever-increasing diffusion of computers in society, human-computer interaction (HCI) is becoming in- creasingly essential to our daily lives. HCI design was dominated first by direct ma- nipulation and then delegation. The tacit assumption of both styles of interaction has been that the human will be explicit, unambiguous, and fully attentive while controlling the information and command flow. Boredom, preoccupation, and stress are unthinkable, even though they are very human behaviors. This insensitivity of current HCI designs is fine for well- codified tasks. It works for making plane reserva- tions, buying and selling stocks, and, as a matter of fact, almost everything we do with computers today. But this kind of categorical computing is inappropri- ate for design, debate, and deliberation. In fact, it is the major impediment to having flexible machines capable of adapting to their users and their level of attention, preferences, moods, and intentions. The ability to detect and understand affective states of a person with whom we are communicating is the core of emotional intelligence. Emotional intelligence (EQ) is a facet of human intelligence that has been argued to be indispensable and even the most important for a successful social life (Goleman, 1995). When it comes to computers, however, not all of them will need emotional intelli- gence, and none will need all of the related skills that we need. Yet man-machine interactive systems capable of sensing stress, inattention, and heedful- ness, and capable of adapting and responding appro- priately to these affective states of the user are likely to be perceived as more natural, more efficacious and more trustworthy. The research area of ma- chine analysis and employment of human affective states to build more natural, flexible HCI goes by a general name of affective computing, introduced first by Picard (1997). BACKGROUND: RESEARCH MOTIVATION Besides the research on natural, flexible HCI, vari- ous research areas and technologies would benefit from efforts to model human perception of affective feedback computationally. For instance, automatic recognition of human affective states is an important research topic for video surveillance as well. Auto- matic assessment of boredom, inattention, and stress will be highly valuable in situations where firm attention to a crucial but perhaps tedious task is essential, such as aircraft control, air traffic control, nuclear power plant surveillance, or simply driving a ground vehicle like a truck, train, or car. An auto- mated tool could provide prompts for better perfor- mance, based on the sensed user’s affective states. Another area that would benefit from efforts toward computer analysis of human affective feed- back is the automatic affect-based indexing of digital visual material. A mechanism for detecting scenes or frames that contain expressions of pain, rage, and fear could provide a valuable tool for violent-con- tent-based indexing of movies, video material, and digital libraries. Other areas where machine tools for analysis of human affective feedback could expand and en- hance research and applications include specialized areas in professional and scientific sectors. Monitor- ing and interpreting affective behavioral cues are important to lawyers, police, and security agents who are often interested in issues concerning decep- tion and attitude. Machine analysis of human affec-