The Roles of Active Perception in Intelligent Agent Systems
14 Jul 2009-Vol. 4078, pp 139-152
TL;DR: The need to incorporate the notion of active perception in intelligent agent systems is discussed and it is argued that active perception plays an imperative role in situation awareness.
Abstract: This paper discusses the need to incorporate the notion of active perception in intelligent agent systems. There are two justifications for this proposition. First, we believe perception should receive more attention and some special treatment. Perception is the most direct and effective way to update an agent's beliefs about the status of its environment. In theory, an overwhelming portion of an agent's decisions about its next course of action, or next actions, in either the reactive or deliberative model, should be heavily influenced by what is happening in the environment --- this is one of the important characteristics that differentiates agents from the traditional software design paradigm. An agent's internal activities, such as deliberation and means-end analysis, to a large extent, are also driven by what have been sensed. The second justification is that active perception plays an imperative role in situation awareness. Active perception is an important capability that directs an agent's limited attention to the relevant aspects of the environment in an proactive manner.
Citations
More filters
22 Aug 2011
TL;DR: In this paper, a framework for perceptual attention for BDI-agents embodied in a virtual environment in the scope of a middleware for connecting a multi-agent system and a game engine is presented.
Abstract: Virtual characters in (serious) games are increasingly required to perform complex tasks in dynamic virtual environments. The use of BDI-agents seems a good fit to realize the intelligent behavior for virtual characters. However, although perception is part of the BDI model, it is not really geared towards real time virtual environments. In this paper we present a framework for perceptual attention for BDI-agents embodied in a virtual environment in the scope of a middleware for connecting a multi-agent system and a game engine. Attention for agents is controlled using an underlying subscription mechanism for sensing the environment. Evaluations are provided based on an implemention of the framework.
16 citations
Dissertation•
01 Jan 2013
TL;DR: The dynamic virtual environment of an IVA is described using entity relationships that abstractly describe the state of the environment, communication actions included in the messages exchanged in chat channels, complex events that take place, and situations that are unfolding.
Abstract: Intelligent Virtual Agents (IVAs) are interactive characters controlled by software agents that exhibit human-like qualities. When interacting with human participants in virtual communities, an IVA is expected to be intelligent, as well as believable. A major factor that determines the intelligence and believability of an IVA is its ability to be aware of its physical and social environments. An important aspect of physical environment awareness is the ability of an agent to refer at an abstract level to the interactions of entities (avatars and objects) in its surrounding environment. However, sensor data received from many virtual worlds describe the agent’s environment at the level of objects and avatars. This has introduced an information representation gap between virtual worlds and IVAs implemented using agent systems such as those based on the BDI architecture. In this thesis, the dynamic virtual environment of an IVA is described using entity relationships that abstractly describe the state of the environment, communication actions included in the messages exchanged in chat channels, complex events that take place, and situations that are unfolding. An abstract model of virtual environments is formulated to describe these different virtual environment elements, and their interrelationships. A set of data inference mechanisms is identified, which interprets the low-level sensor data received from virtual worlds into the aforementioned elements of dynamic virtual environments. A framework implementing these data inference mechanisms is used to process sensor data received by Jason agents deployed in the popular multi-purpose virtual world Second Life.
6 citations
Cites background or methods from "The Roles of Active Perception in I..."
...Construction of a situation model is the main prerequisite for real-time situation recognition in complex dynamic systems (Buford et al., 2006; So and Sonenberg, 2009)....
[...]
...Top-down filtering refers to retaining only those percepts that are relevant to the currently pursued goals of the agent (So and Sonenberg, 2009)....
[...]
...In contrast, So and Sonenberg (2009) identify a situation as “a narrative of observable events, which lead to a particular state of the environment when they all happened” (So and Sonenberg, 2009)....
[...]
...In the BDI model, situations are normally treated as beliefs (Buford et al., 2006; So and Sonenberg, 2009)....
[...]
02 Jun 2015
TL;DR: A design for a percept buffer that, when configured with domain-specific percept metadata and application- specific percept management policies, provides a generic but customisable solution to the task of developing agents that perceive information from asynchronously executing external systems, especially those producing data at a high frequency.
Abstract: Multi-agent systems technologies have been widely investigated as a promising approach for modelling and building distributed systems. However, the benefits of agents are not restricted to systems solely comprised of agents. This paper considers how to ease the task of developing agents that perceive information from asynchronously executing external systems, especially those producing data at a high frequency. It presents a design for a percept buffer that, when configured with domain-specific percept metadata and application-specific percept management policies, provides a generic but customisable solution. Three application case studies are presented to illustrate and evaluate the approach.
4 citations
Cites background from "The Roles of Active Perception in I..."
...Top-down filtering refers to retaining only those percepts that are relevant to the currently pursued goals of the agent [14]....
[...]
Dissertation•
20 Jun 2013
TL;DR: In this article, the authors propose a double activite de perception (passive and active), couplee a limitation quantitative des percepts (limite de la memorisation a court-terme), and teste dans le contexte d'un carrefour.
Abstract: Dans le domaine de la modelisation du trafic routier, des ameliorations sont necessaires afin de permettre de simuler des situations accidentogenes. La grande majorite des modeles microscopiques actuels a base d'agents pour la simulation de trafic concerne l'activite decisionnelle du conducteur. Or, l'activite de perception est un prealable important a la decision. Notre objectif est donc de proposer de nouvelles solutions pour la modelisation agent de la perception du conducteur humain. La proposition comprend une double activite de perception (passive et active), couplee a une limitation quantitative des percepts (limite de la memorisation a court-terme). Le modele a ete implemente et teste dans le contexte d'un carrefour pour lequel nous possedons un ensemble de donnees d'observation. Les resultats obtenus confirment l'interet du modele pour simuler des comportements proches des comportements observes et des variations inter-individuelles dans la population de conducteurs.
4 citations
04 Jun 2012
TL;DR: The computational complexity of the sensing process in the game engine can be controlled by an agent itself, allowing it to deal with more complex environments, and an overall performance analysis of the framework is provided.
Abstract: When agents are integrated in a game engine for embodiment in a virtual environment, perception often leads to performance issues due to the lack of control over the sensing process In previous work a perception framework was proposed within CIGA, a middleware facilitating the coupling between a multiagent system and a game engine It allowed agents to have control over the flow of sensory information generated in their embodiment In this paper we continue this work by presenting performance optimizations within this framework Here, the computational complexity of the sensing process in the game engine can be controlled by an agent itself, allowing it to deal with more complex environments Additionally we provide an overall performance analysis of the framework
3 citations
Cites background from "The Roles of Active Perception in I..."
...So and Sonenberg [14] relate their work with Weyns et al....
[...]
References
More filters
TL;DR: A theoretical model of situation awareness based on its role in dynamic human decision making in a variety of domains is presented and design implications for enhancing operator situation awareness and future directions for situation awareness research are explored.
Abstract: This paper presents a theoretical model of situation awareness based on its role in dynamic human decision making in a variety of domains. Situation awareness is presented as a predominant concern in system operation, based on a descriptive view of decision making. The relationship between situation awareness and numerous individual and environmental factors is explored. Among these factors, attention and working memory are presented as critical factors limiting operators from acquiring and interpreting information from the environment to form situation awareness, and mental models and goal-directed behavior are hypothesized as important mechanisms for overcoming these limits. The impact of design features, workload, stress, system complexity, and automation on operator situation awareness is addressed, and a taxonomy of errors in situation awareness is introduced, based on the model presented. The model is used to generate design implications for enhancing operator situation awareness and future directio...
7,470 citations
Book•
01 Jan 1958TL;DR: In this paper, the authors describe a transition between behaviourist learning theory and the modern information processing or cognitive approach to perception and communication skills, and provide a principal starting point for theoretical and experimental work on selective attention.
Abstract: First published in 1958, this book has become recognized as a classic in its field. It marked a transition between behaviourist learning theory and the modern 'information processing' or 'cognitive' approach to perception and communication skills. It continues to provide a principal starting point for theoretical and experimental work on selective attention. As Professor Posner writes in his Foreword to the reissue: 'it remains of great interest to view the work in its original form and to ponder those creative moments when the mind first grasps a new insight and then struggles to work out its consequences.
5,325 citations
Book•
12 Jun 2002
TL;DR: A multi-agent system is a distributed computing system with autonomous interacting intelligent agents that coordinate their actions so as to achieve its goal(s) jointly or competitively.
Abstract: The study of multi-agent systems (MAS) focuses on systems in which many intelligent agents interact with each other. These agents are considered to be autonomous entities such as software programs or robots. Their interactions can either be cooperative (for example as in an ant colony) or selfish (as in a free market economy). This book assumes only basic knowledge of algorithms and discrete maths, both of which are taught as standard in the first or second year of computer science degree programmes. A basic knowledge of artificial intelligence would useful to help understand some of the issues, but is not essential. The books main aims are: To introduce the student to the concept of agents and multi-agent systems, and the main applications for which they are appropriate To introduce the main issues surrounding the design of intelligent agents To introduce the main issues surrounding the design of a multi-agent society To introduce a number of typical applications for agent technology
4,042 citations
Book•
01 Nov 2001
TL;DR: A multi-agent system (MAS) as discussed by the authors is a distributed computing system with autonomous interacting intelligent agents that coordinate their actions so as to achieve its goal(s) jointly or competitively.
Abstract: From the Publisher:
An agent is an entity with domain knowledge, goals and actions. Multi-agent systems are a set of agents which
interact in a common environment. Multi-agent systems deal with the construction of complex systems involving multiple agents and their coordination. A multi-agent system (MAS) is a distributed computing system with
autonomous interacting intelligent agents that coordinate their actions so as to achieve its goal(s) jointly or competitively.
3,003 citations