Affective Interactions for in Real-time Applications: the SAFIRA Project
01/2004
by Yasmine Arafa, Luís Miguel Botelho, Adrian Bullock, Pedro Figueiredo, Patrick Gebhard, Kristina Höök, E. H. Mamdani, Ana Paiva, Paolo Petta, Phoebe Sengers, Marco Vala
This paper provides an overview of the SAFIRA project, an IST-KA4 project in the area of Affective Interactions. We explain the rationale for providing computers with the ability to understand, model, and express human-like emotions. We present the toolkit for affective computing that we are building, and describe 3 demonstrators that illustrate different approaches to affective interaction that are using our toolkit.
view more
Emotion Eliciting in Salt & Pepper
01/2004
by Luis Botelho, Pedro Ramos, Pedro Figueiredo
This paper describes the emotion elicitation process of the Salt & Pepper architecture for autonomous agents. Its main contributions are a general framework upon which it is possible to define different theories of emotion generation, the view of artificial emotion as an adaptive mechanism that is designed at the level of the agent architecture, and use and the discussion of several emotion eliciting mechanisms with special emphasis on non-cognitive emotion generation processes. We propose to add emotion eliciting inhibition times to emotion eliciting rules in order to avoid the repeated generation of the same emotion due to the same circumstances. Finally, we propose the conditioned emotion eliciting chunk, which sequentially generates conditioned emotions. This new emotion eliciting structure is useful to initiate emotion processes in stereotypical situations in which the initiated emotion depends on the emotion that has previously been generated. Our experiments were done during the SAFIRA European Project.
view more
Send Fredo off to do this, send Fredo off to do that
07/2003
by Lus Botelho, Hugo Mendes, Pedro Figueiredo, Rui Marinheiro
Fredo is a generic domain-independent broker that creates value- added information taking into account the preferences specified by its clients. Fredo uses ontology services and yellow pages services to discover a set of agents that can provide information relevant to its clients' requests. Fredo uses an intelligent heuristic strategy based on a fuzzy evaluation mechanism to plan the queries it uses to gather relevant information for its clients' needs. In order to handle possible information overload, we have designed a special purpose interaction protocol, the paged information-request protocol, which is used to govern the interaction between Fredo and information providers. Fredo also uses a fuzzy inference engine to evaluate the gathered information with respect to the preferences specified by its clients. Fredo has been developed by and used in the Agentcities project. Fredo uses the FIPA ACL inter agent commu- nication language with FIPA SL contents. It was implemented in JAVA and Prolog and runs on FIPA++, a FIPA compliant agent platform.
view more
Video-based multi-agent traffic surveillance system
02/2000
by B. Abreu, L. Botelho, A. Cavallaro, D. Douxchamps, T. Ebrahimi, P. Figueiredo, B. Macq, B. Mory, L. Nunes, J. Orri, M.J. Trigueiros, A. Violante
This paper describes Monitorix, a video-based traffic surveillance multi-agent system. Monitorix agents are grouped in four tiers, according to the kind of information processing they perform: the sensors and effectors tier, the objective description tier, the application assistant tier, and the user assistant tier. The video analysis algorithms use an adaptive, data-driven, application independent approach to extract features from the video raw data. In spite of the diversity of agent tasks, adaptive learning algorithms are used in most cases. The integration of video analysis algorithms and agent technology is made via a special middle agent called Proxy. Monitorix is a fully decentralised multi-agent system living in a FIPA Platform and using FIPA Agent Communication Language. Tracking of vehicles across nonoverlapping cameras is performed by the Tracker agent, using a traffic model and learning algorithms that tune the model parameters.
view more