Attending to attributes in double-conjunction texture segregation: The role of color, luminance, and orientation

Purpose. To determine whether certain attributes can be primed for selective distributed attention in the texture segregation process. Methods. We used texture edges defined by double conjunctions (DCJ). The left panel shows a DCJ of orientation (O) and luminance-polarity (P). Edge extraction can be thought of proceeding in two stages: 1) Grouping in two subsets: the subset of all dark texels of the left panel is in the middle panel; the subset of all 45°-oriented texels is in the right panel. 2) Texture segregation within a grouped subset: this is done on the basis of O in the middle panel; on the basis of P in the right panel. 18 observers were shown DCJ of P and O (left panel), and DCJ of coJor (C) and O (same as left panel with bright and dark texels replaced by equiluminant red and green, respectively). They were asked to attend to one attribute, and report the orientation of the global texture edge (horizontal or vertical) with a stimulus duration of 833 ms in a 2AFC paradigm. Results. a) 3 observers were not able to detect the edge no matter which attribute they attended to. b) For the rest, color is slightly easier to attend to than polarity (40% contrast). Orientation was almost as easy to attend to as C or P. Conclusions. The results indicate minor differences on how attributes can be primed in a distributed attentional mode, which may be linked to their ability to group subsets. (Figure Presented).