Catálogo de publicaciones - libros

In recent years, emotional computing has found an important application domain in the field of interactive synthetic characters. Interesting examples of this domain are computer games, interface agents, human-robot interaction, etc. However, few systems in this area include a model of personality, although it plays an important role in differentiating agents and determining the way they experience emotions and the way they behave.

In order to synthetize affective states, computer systems use emotion theories from psychology, that provide guidelines for deciding what kind of emotion, and with what intensity, should be elicited in a given situation. According to most theories, the emotions of an agent depend on the subjective appraisal of situations. While they provide essential knowledge for the generation of synthetic emotions, these theories do not specify all aspects necessary for the design of computational systems that implement them. So, when designing computational agents that “experience” affective states, computer scientists have a long way to go, and many decisions to make, to produce the information necessary to theories of emotion. When all these difficulties are overcome, still a lot of experimenting is necessary to make adjustments to the formulas and algorithms used.

In recent times several attempts have been made to give a robot or broader spoken a computer some kind of feelings in order to understand and model human capacities. The main idea of our work was the design of expressive robot movements for the display of emotional content embedded in the audio layer in both live and recorded music performance. Starting from results in studies on musicians’ body in emotional expressive music performance (see [3]), we tried to map different movement cues (e.g. speed, fluency) to movements of a small mobile robot. The robot had constraints of sensors and motors, so the emotions were implemented taking into account only the main characteristics of musicians’ movements. We implemented movements for the three emotions happiness, anger and sadness. Subjects were asked to judge in a perceptual test which emotional intentions were communicated by the movements.

This work aims at measuring the anticipated perception of emotions on minimal linguistic units, to evaluate if the underlying cognitive processing is compatible with the hypothesis of gradient contours. Selected monosyllabic stimuli extracted from an expressive corpus and expressing anxiety, disappointment, disgust, disquiet, joy, resignation, sadness and satisfaction, were gradually presented to naïve judges in a gating experiment. Results strengthen the hypothesis of gradient processing by showing that identification along successive gates of most of expressions follow a linear pattern typical of a contour-like processing, while expressions of satisfaction present distinct gradient values that make possible an early identification of affective values.

By observing the way people who are stressed interact with objects one may be able to interpret how they feel (Krauss et al., 1996). For instance, at a presentation a nervous presenter can often be seen fidgeting with a pen or pointer. This has shown to be true even if people are actively trying to suppress or hide these feelings (Ekman & Friesen, 1967). The behaviors when manipulating objects during a stressful event appear to be qualitatively different from manipulations during boring or neutral events (Kenner, 1984). Yet, no studies were found in which a categorization was made in the types of manipulative hand movements during different arousing and non-arousing conditions. The goal of this study was to explore what manipulative hand movements are evoked when using a pen during a stressful event as compared to a neutral event.

This paper discusses a data driven pilot study, designed to explore the expression of emotion in a natural situation.Short sequences of natural behaviour were transformed into still photos and compared with examples of acted behaviour using a FeelTrace type analysis. Results indicate that rapid transitions are more common in natural than in acted behaviour.

Human synthetic characters in any computer simulated environment need to be capable of behaving the appropriate way of expressing emotion like normal human do. If the characters fail to express the suitable emotional expression, they will likely to break the users mood and belief. These imply that the key problem in designing synthetic characters is to make them believable in their overall behavior. Therefore the study was directed to construct human synthetic characters that can improve humans respond as well as identify intensity rating of emotions. Ekmans six emotional expressions [1], which are happiness, sadness, anger, surprise, fear and disgust were used because these basic emotions are clear, widely accepted and sufficient. An experiment was conducted to measure believable emotional expression from static images. Sets of virtual human static images of posed expressions of emotions were shown to the subjects. A total of 36 thirty-six volunteers (18 men and 18 women) took part in the experiment. The mean age was 29 years old. The subjects were asked to recognize the expressions by grouping them into six basic emotions and provide the level of emotion for each posture to study the intensity-rating tasks.

The artwork (2003), by Tenhaaf with sound by Kamevaar, has a pod-like amorphous shape and affords the feeling to people that they are dealing with an entity. It has been dubbed by some interactants as a “baby” robot, despite the fact that it has no moving parts and only one ultrasonic distance sensor for detecting its environment (and thus is not robotic). But it has ”baby talk” suitable for a machine: sound as pure signal, that could have no other origin than electronic signal flow itself. Interactants set off electronically-manipulated microphone feedback sounds when proximal to - sounds that become louder and more intense when the interactant moves away and softer as she or he comes closer. The generated sounds are layered: several sounds playing at once generate the assault of noise, whereas a single sound is almost melodic. Through its sound, both commands its space and directs people’s movements; it is thereby perceived as having the potential to mature into a more autonomous entity. The idea of the work is not to elicit beliefs that one is seeing life emerging artificially, but rather to elicit a willingness to talk to this entity.

I-Sounds aims to increase the Affective Bandwidth of an Interactive Drama system called I-Shadows, implementing a fully emergent system that generates affective music, based on musical theory and on the emotional state of the characters.

is formulated as a specific sub-problem of cognitive appraisal. A computational model of goal appraisal is built via a knowledge-based system. An empirical validation study compared model versus human performance by eliciting believability judgments from naive human raters. Results showed that computer model goal appraisals were, in most cases, as believable as human-generated appraisals.