Purchase this article with an account.
Sofia Lavrenteva, Ikuya Murakami; Perceiving and grasping the equiluminant Ebbinghaus illusion. Journal of Vision 2019;19(10):110a. doi: https://doi.org/10.1167/19.10.110a.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
According to two streams hypothesis, (1) actions rely on the dorsal while perception on the ventral stream, (2) These streams process shapes within different frames of reference (egocentric vs. relative), and (3) the dorsal stream is less affected by the Ebbinghaus illusion, in which the perceived size of a central target is affected by the sizes of surrounding inducers. However, many studies indicate equivalent illusion strength between perception and action, supporting single shape representation used in both streams. We attempted to create inducer disks invisible to the dorsal stream, taking advantage of the tendency that it mainly gets its input through the magnocellular pathway insensitive to S-cone modulation. We measured how the illusion strength differed between perception and grasping in stimuli comprised of luminance defined and S-cone modulating equiluminant inducer disks. We used graspable disks (2-mm thick transparent plastic) of 3 different sizes as target disks. In each trial, the target disk was surrounded by either large or small inducer disks, defined either by luminance or by S axis in DKL color space. In a perception test, observers adjusted the size of a reference disk to match the apparent size. In an action test, they grasped the target disk with a precision grip. We only analyzed the data of the observers whose maximum grip aperture (MGA) was significantly affected by physical disk size. In the perception test, the Ebbinghaus illusion was similar in strength between the stimuli with luminance defined and equiluminant inducers. In the action test, however, the effect of the Ebbinghaus illusion on the MGA was stronger in the stimulus with the luminance inducers than in the stimulus with the equiluminant inducers. These results neither fully support different frames of reference in the two streams, nor a single shape representation. We discuss possible interactions between the two streams.
This PDF is available to Subscribers Only