Simulation and Formal Analysis of Visual Attention in Cognitive Systems
In this paper a simulation model for visual attention is discussed and formally analysed. The model is part of the design of a cognitive system which comprises an agent that supports a naval officer in its task to compile a tactical picture of the situati
- PDF / 744,729 Bytes
- 18 Pages / 430 x 660 pts Page_size
- 35 Downloads / 224 Views
Vrije Universiteit Amsterdam, Department of Artificial Intelligence De Boelelaan 1081a, 1081 HV Amsterdam, The Netherlands [email protected], {tbosse, treur}@cs.vu.nl 2 TNO Human Factors, P.O. Box 23, 3769 ZG Soesterberg, the Netherlands
Abstract. In this paper a simulation model for visual attention is discussed and formally analysed. The model is part of the design of a cognitive system which comprises an agent that supports a naval officer in its task to compile a tactical picture of the situation in the field. A case study is described in which the model is used to simulate a human subject’s attention. The formal analysis is based on temporal relational specifications for attentional states and for different stages of attentional processes. The model has been automatically verified against these specifications.
1 Introduction The model of visual attention discussed and formally analysed in this paper is part of the design of a cognitive system which comprises an agent that supports a naval officer in its task to compile a tactical picture of the situation in the field. In this domain, the complex and dynamic nature of the environment makes that the officer has to deal with a large number of tasks in parallel. Therefore, in practice (s)he is often supported by agents that take over part of these tasks. However, a problem is how to determine an appropriate work division: due to the rapidly changing environment, such a work division cannot be fixed beforehand [2]. This results in a need for reallocation of work which is determined dynamically and at runtime. For this purpose, two approaches exist, namely human-triggered and system-triggered dynamic task allocation [8]. In the former case, the user can decide up to what level the system (or agent) should assist him. But especially in alarming situations the user does not have enough time to think about task reallocations [18]. In these situations it would be better if the system determines this. Hence a system-triggered dynamic task allocation is desirable. In order to obtain such a system-triggered dynamic task allocation, the model of visual attention discussed and formally analysed in this paper can be incorporated within the supporting agent. The idea is to use an estimation of the user’s current ∗
Parts of this paper are based on work presented at the 2006 IEEE/WIC/ACM International Conference on Intelligent Agent Technology (IAT’06) [5] and the Second European Cognitive Science Conference (EuroCogSci'07) [6].
L. Paletta and E. Rome (Eds.): WAPCV 2007, LNAI 4840, pp. 463–480, 2007. © Springer-Verlag Berlin Heidelberg 2007
464
T. Bosse, P.-P. van Maanen, and J. Treur
attention to determine which subtasks the agent is best to pay attention to. For instance, if the user has the subtask to pay attention to a certain track on the screen, it is a possibility is that no additional support for that track is needed. In this case the agent should rather direct its own ‘attention’ to the user’s unattended tracks. The assumption made here, that the allocat
Data Loading...
