Course decsription
Tamagotchis , Aibo pets and the virtual inhabitants of the computer game The Sims can be seen as "artificial beings" with at least one important thing in common; they aim to interact with humans in a natural way. One way to achieve this, most obvious in Aibos and The Sims, is to embed a Computational Model of Emotion in these artificial beings. The field of computing with emotion is called Affective Computing. Even the simple emotions of the Aibo dog and the inhabitants of the computer game The Sims are based upon theories of emotion. In this course you will learn to recognize different views on emotion, as well as how and why these views can be (have been / should be) used to develop computational models of emotion. We will discuss several existing models. Finally we will together develop a new, complete system that uses many of the affective computing elements discussed during the course. Overall, this course will give you insights into the fun (gaming, robots) and glory (theoretical predictions) of affective computing.
Main navigation menu
Course
Examination.
Class participation (33%). A paper on a topic within the field of affective computing (33%). Practical course work; as a group we will develop a complete emotional system (33%).
Course organization.
When: During summer break from July-1st till July 28th. 4 Weeks, full time (or nearly, that is, 32 hours per week, as some people might have holiday jobs). Week 1,2: 6 lectures of 2 hours each. You are required to study material in between the lectures. Weeks 3,4: practical course work. Where: all activities are at the LIACS (Snellius building). There is a maximum of 12 participants.
Prerequisites.
Intermediate to advanced programming experience in C(++)/Java/Flash script (or other OO language), and completion of at least one of the following courses: Multimedia Programming, Multimedia Systems or Human Computer Interaction (however, this last condition is negotiable if you are sufficiently motivated and at least a 4rd year student).
Information.
Joost Broekens, broekens@liacs.nl, Room 120, Phone: 071-5275779.
Pascal Haazebroek, pascal@haazebroek.nl
Enrollment
As there is a maximum of 12 participants, please consider the following. If you are enthusiastic about participating in this new and exciting course, please register before 1st of May by sending an email to either Riet Derogee (derogee@liacs.nl) if you are a computer science student, or Martine Roeleveld (roelevel@liacs.nl) if you are a media technology student. We hope many of you will show up!
Notes/Announcements
ATTENTION: course enrollment is possible until May 31-th!! So if you're planning to do the course, please hurry signing up.
Contact information / TopIntro
Tamagotchis, Aibo pets and the virtual inhabitants of the computer game The Sims can be seen as "artificial beings" with at least one important thing in common; they aim to interact with humans in a natural way. One way to achieve this, most obvious in Aibos and The Sims, is to embed a Computational Model of Emotion in these artificial beings.
Such model enables the artificial being to simulate and show emotions that are interpretable by us humans. In the context of these artificial beings, specifically when the being is a "virtual pet", the main aim of a computational model of emotion is to achieve stronger attachment between human and pet.
Several researchers are trying to bring this aspect of emotional interaction between artificial beings and humans to a higher level. They embed computational models of emotion in robots to study human robot interaction (HRI), or to study how emotional communication can be used to enhance learning abilities of robots (affective learning). Obviously, apart from the need to simulate and show emotions, these models also need to recognize human emotions.
In order to recognize and act upon an emotion, the model must interpret the emotion itself. What does it mean to be happy, or sad? Cognitive scientists and artificial intelligence researchers use computational models of emotion to address exactly this question. What does a certain emotion mean in terms of beliefs about the world, and what are the potential effects of that emotion? For example, why am I happy when I win the lottery? Computational models of emotion can help address these questions within a cognitive framework as they concretize cognitive theories. When embedded in an artificial being, such model shows emotional reactions to situations. As these models are based upon a theory, it is in fact the theory that explicitly predicts a certain emotion. Such models can thus be used to advance the theory of emotion.
Even the simple emotions of the Aibo dog and the inhabitants of the computer game The Sims are based upon theories of emotion. In this course you will learn to recognize different views on emotion, as well as how and why these views can be (have been / should be) used to develop computational models of emotion. We will discuss several existing models. Finally we will together develop a new, complete system that uses many of the affective computing elements discussed during the course. Overall, this course will give you insights into the fun (gaming, robots) and glory (theoretical predictions) of affective computing.