Abstract
Visual features are often assumed to be the general building blocks for various visual tasks. However, it is well known that some stimulus categories (i.e., basic features) can be processed in parallel, but others (e.g., Ts in different orientations) need to be scanned serially, and this difference in featural strength seems to be on a fundamentally different dimension from differences in visual strength (e.g., reduction in contrast). This study compared two high-level tasks, namely tasks that require a lot of attentional operations (change detection and pattern comparison), with one low-level task, namely a task that requires few attentional operations (perceptual discrimination). The results confirmed that featural strength has substantial effects on high-level tasks but only a negligible effect on the low-level task. The results also revealed a complementary interaction: Visual strength has a substantial effect on the low-level task, but a negligible effect on high-level tasks. Overall, featural strength and visual strength are two dissociable dimensions in processing of visual features. The present results, along with other findings, challenge the generality of processing visual features.
Highlights
Visual features are often assumed to be the general building blocks for various visual tasks
The single-dimension notion—namely that the featural strength and visual strength boil down to a single general strength—will only be supported if we find that visual strength has a substantial effect on a high-level task, but a negligible effect on a low-level task
The results indicate that featural strength has substantial effects in high-level tasks, but not in low-level tasks, whereas the visual strength of features has substantial effects in low-level tasks, but not in high-level tasks
Summary
Visual features are often assumed to be the general building blocks for various visual tasks. A previous study found that across a range of features, the visual working memory capacities and search slopes for these features are perfectly correlated[1]. This seems to imply that performance on these tasks, when measured in appropriate ways, is exactly determined by a single dimension of general strength of featural differences. The term “visual features” refers to both basic features (e.g., colors, shapes) and non-basic features (e.g., Ts in different orientations) Consistent with this use, this previous study[1] included both basic features and non-basic features, and their results indicated that the relationship found between working memory capacities and search slopes applies to both. Feature integration theory[3], for example, proposed an parallel stage, in which a set of basic features (e.g., color, shape, size) are extracted in parallel, followed by a serial stage, in which attention allows the binding of these features so that they can be consciously perceived as a whole
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
