Abstract
Aspect-ratio (height:width ratio) is a fundamental property for shape discrimination. Aspect-ratio might be detected by combining separate estimates of height and width, or by a specialized aspect-ratio detector sensitive to height:width ratios independently from separate linear dimensions. Previous work suggests aspect ratio is the primary cue. The current study further tested these two hypotheses for aspect-ratio detection by measuring size (2D area) aftereffects and aspect-ratio aftereffects. In a novel procedure, employing forced-choice psychophysical methods, the size aftereffects were used to predict an aspect-ratio aftereffect consistent with the height-and-width-combination hypothesis, which was opposite in direction to that predicted by the specialized aspect-ratio detector hypothesis. This was possible because a preliminary experiment, using square stimuli, showed size aftereffects exhibit a non-monotonic relationship to adaptor:test size ratios. The results showed that the direction of the aspect-ratio aftereffect was consistent with the specialized aspect-ratio detector hypothesis. In an extension of this study aspect-ratio aftereffects were then measured, for a large range of adaptor:test ratios, to determine whether this specialized detector represents aspect-ratio using a multi-channel or opponent coding system. The results showed smaller after effects with large adaptor:test ratios than with intermediate ratios. This outcome is consistent with multi-channel coding, i.e. that aspect-ratio is detected by multiple mechanisms sensitive to small, overlapping ranges of aspect-ratio.
Meeting abstract presented at VSS 2014