Purchase this article with an account.
Karla Evans, Lucy Spencer; Allocation of Attention in a Complex Environment. Journal of Vision 2017;17(10):753. doi: 10.1167/17.10.753.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Visual attention is a set of control mechanisms that adapt the visual system for completion of different perceptual tasks, allowing us to perceive our complex environment. There are two prominent views on how allocation of this resource affords us the ability to individuate and recognize an object in a complex scene, and also register global properties of the same scene. One view sees this as workings of one single process whose activity ranges from narrowly focused analysis of local binding of features, to global registration of image and summary statistics (Tresiman, 2006). The opposing view argues that it is a result of two processes (one selective and one non-selective), but does not make a prediction of whether they work in parallel or not (Wolfe et al., 2011). Does perception operate along a continuum of one process, or using two processes in serial or in parallel? We used a dual-task paradigm to test these three possible models, in which a task requiring global processing is performed simultaneously with another task requiring the same, or with a task requiring focused processing. Observers completed three single-task conditions (image categorization, average global dot movement, and central single dot movement). The difficulty on the two dot tasks were equated. This was followed by two dual-task conditions, one of which used two global tasks (image categorization and average global dot movement), the other of which used both a global and a focused task (image categorization with single dot motion direction). We compared the observers' performance while they performed two tasks simultaneously to their performance on single-task conditions. The results are in favour of a single process with both global and single dot tasks showing the same degree of reduction in accuracy during dual tasks, but with no reduction in image categorization accuracy.
Meeting abstract presented at VSS 2017
This PDF is available to Subscribers Only