11:15   Oral Session 5-KzZ – Emotion Evaluation 2
Chair: Egon L. van den Broek
11:15
25 mins 		Emotion Attribution to Basic Parametric Static and Dynamic Stimuli
Valentijn Visch, Martijn Goudbeek
Abstract: The following research investigates the effect of basic visual stimuli on the attribution of basic emotions by the viewer. In an empirical study (N = 33) we used two groups of minimal expressive stimuli: dynamic and static. The dynamic stimuli consisted of an animated circle moving according to a structured set of movement parameters, derived from emotion expression literature. The parameters are direction, expansion, velocity variation, fluency, and corner bending. The static stimuli consisted of the minimal visual form of a smiley. The varied parameters were mouth openness, mouth curvature, and eye rotation. The findings describing the effect of the parameters for an attributed emotion are presented. This paper shows how specific viewer affect attribution can be included in men machine interaction using minimal visual material.
11:40
25 mins 		The challenges of dealing with distributed signs of emotion: theory and empirical evidence
Roddy Cowie, Ellen Douglas-Cowie, Gary McKeown, Ceire Gibney
Abstract: There clearly are important issues associated with the distinction between acted and naturalistic data, but focusing on acting may not be the best way to articulate them. An alternative is to focus on differences of structure which are often (but not always) associated with the distinction. Several such differences relate to the way signs are distributed: simultaneous or sequential appearance; in one modality or many; stable or unstable; suggesting one emotion or many; briefly or continually available. These distinctions clearly relate to different demands on analysis.
12:05
25 mins 		The Importance of the Body in Affect-Modulated Action Selection: A Case Study Comparing Proximal versus Distal Perception in a Prey-Predator Scenario
Claire Suzanne O'Bryne, Lola Cañamero, John C. Murray
Abstract: In the context of the animat approach, we investigate the effect of an emotion-like hormonal mechanism, as a modulator of perception~---~and second order controller to an underlying motivation-based action selection architecture~---~on brain-body-environment interactions within a prey-predator scenario. We are particularly interested in the effects that affective modulation of different perceptual capabilities has on the dynamics of interactions between predator and prey, as part of a broader study of the adaptive value of emotional states such as ``fear'' and ``aggression'' in the context of action selection. In this paper we present experiments where we modulated the architecture of a prey robot using two different types of sensory capabilities, proximal and distal, effectively creating combinations of different prey ``brains'' and ``bodies''.