Catálogo de publicaciones - libros

Executable computational process models of emotion are based on specific sets of modelling primitives. Motivated by the requirements of a specific scenario and concepts used by emotion theories, we propose as building blocks explicitly bounded resources and concurrent processes acquiring and using them. Our approach is intended for the incremental modelling of a growing collection of emotional episodes, with a clear delineation of technically necessary simplifications of the natural phenomena. An episode of disgust is used to discuss the approach, which is realised using real-time cooperative microthreading technology.

Artificial agents engaged in real world applications require accurate resource allocation strategies. For instance, open systems may require artificial agents with the capability to filter out all information which are irrelevant with respect to the actual intentions and goals. In this work we develop a model of surprise-driven belief update. We formally define a strategy for epistemic reasoning of a -inspired agent, where surprise is the causal precursor of a belief update process. According to this strategy, an agent should update his beliefs only with inputs which are surprising and relevant with respect to his current intentions. We also compare in practice the performances of agents using a surprise-driven strategy of belief update and agents using traditional reasoning processes.

We analyze the role of motivations in living organisms, and the nature of their influences on behavior with the aim to propose a design methodology for schema-based agent architectures. We propose that motivations have a modulatory influence on behavior, and in our design methodology they regulate the allocation of resources to the sensorimotor system and schemas. We describe an agent architecture incorporating this principle and we highlight its performance in a simulative study.

This paper outlines the design of a model amalgamating computational visual attention and emotion approaches for the purposes of driving expressive attentive and emotional behaviour in an embodied, situated real-time virtual infant. The themes of emotion and attention underlie all aspects of our model: from perception, to memory, internal state and behaviour expressivity. The model is focused on some of the earliest stimulus evaluation checks related to appraisal theory: perceptual attention focuses and refines the details of relevant and potentially relevant stimuli, reducing uncertainty in order to discover reward and heed danger. In the process, the agents internal state is modified and feeds back to modulate the ongoing allocation of attention and processing of stimuli. Changes in internal state are expressed through a repertoire of prototypical infant gaze, face and body behaviours. This represents, to our knowledge, the first attempt to marry all of these concepts in a real-time 3D embodied agent system.

In this paper, we take an incremental, bottom-up approach to investigate plausible mechanisms underlying emotional modulation of behavior selection and their adaptive value in autonomous robots. We focus in particular on achieving adaptive behavior selection in competitive robotic scenarios through modulation of perception, drawing on the notion of biological hormones. We discuss results from testing our architectures in two different competitive robotic scenarios.

The paper presents a conceptual analysis of enthusiasm in terms of a goal and belief model of emotions. Enthusiasm is an emotion of the same family of joy, felt when an Agent believes she will very likely achieve a goal she is pursuing, because she has the necessary internal resources to achieve it. The function of enthusiasm is to enhance energy and persistence in goal pursuit. Some positive and negative aspects of the contagion of enthusiasm are presented and some empirical studies about enthusiasm and its contagion briefly overviewed.

To build autonomous robots able to live and interact with humans in a real-world dynamic and uncertain environment, the design of architectures permitting robots to develop attachment bonds to humans and use them to build their own model of the world is a promising avenue, not only to improve human-robot interaction and adaptation to the environment, but also as a way to develop further cognitive and emotional capabilities. In this paper we present a neural architecture to enable a robot to develop an attachment bond with a person or an object, and to discover the correct sensorimotor associations to maintain a desired affective state of well-being using a minimum amount of prior knowledge about the possible interactions with this object.

This research focuses on designing affective roles in agent-based social simulation (ABSS) focused on ethology. Synthetic agents are addressed as autonomous, intentional software entities capable of managing primate-like (hierarchical) social relationships in small-scale societies. The critique involves discussion of potential affective roles in socio-cognitive agent architectures, both in terms of individual action-selection and group organisation. With the diversity of social and emotional accounts, primate-like ABSS is put forward with individual behaviour related not only to reactivity or focused on function-optimisation.

Working with emotion-related states in technological contexts requires a standard representation format. Based on that premise, the W3C Emotion Incubator group was created to lay the foundations for such a standard. The paper reports on two results of the group’s work: a collection of use cases, and the resulting requirements. We compiled a rich collection of use cases, and grouped them into three types: data annotation, emotion recognition, and generation of emotion-related behaviour. Out of these, a structured set of requirements was distilled. It comprises the representation of the emotion-related state itself, some meta-information about that representation, various kinds of links to the “rest of the world”, and several kinds of global metadata. We summarise the work, and provide pointers to the working documents containing full details.

Human-to-computer interaction in a variety of applications could benefit if systems could accurately analyze and respond to their users’ affect. Although a great deal of research has been conducted on affect recognition, very little of this work has considered what is the appropriate information to extract in specific situations. Towards understanding how specific applications such as affective tutoring and affective entertainment could benefit, we present two experiments. In the first experiment, we found that students’ facial expressions, together with their body actions, gave little information about their internal emotion per se but they would be useful features for predicting their self-reported “true” mental state. In the second experiment, we found significant differences between the facial expressions and self-reported affective state of viewers watching a movie sequence. Our results suggest that the noisy relationship between observable gestures and underlying affect must be accounted for when designing affective computing applications.