Purchase this article with an account.
Daniel Linares, Isamu Motoyoshi, Kazushi Maruya, Shin'ya Nishida; Motion attached to a new surface is easier to detect. Journal of Vision 2011;11(11):1040. doi: 10.1167/11.11.1040.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The suprathreshold appearance of attributes such as color and motion depends not only on the early processing of specialized detectors, but also on how the image is segmented into surfaces. Detection thresholds, however, are thought to primarily depend on the activity of early detectors. Contrary to this view, here we show that motion detection sensitivity is enhanced when motion is perceived as belonging to a new surface. Base line condition: a field of random dots, consisting of dots moving to the left or to the right and dynamic noise dots, was displayed for 133 ms. This signal period was preceded and followed by 500 ms periods of dynamic noise dots. A tone was presented at the commencement of the signal period to minimise temporal uncertainty about the signal. New surface condition: during the signal period, four pac-men, arranged to form an illusory Kanizsa square, suddenly appeared inside the dot field. The illusory figure gave the impression that the dots it enclosed were on the new square-shaped surface. RESULTS: Motion coherence thresholds were much lower for the new surface relative to the base line condition. This sensitivity improvement did not occur if the pac-men were presented throughout, indicating that the effect is caused neither by the pac-men providing a stationary cue that facilitates the perception of relative motion, nor due to a change in size of the random dots area. Furthermore, sensitivity did not improve when visual cues not resulting in perception of a new surface were displayed during the signal period (for example, only one pacman), implying that the enhancement is not simply caused by exogenous attentional cueing. The transient Kanizsa surface may temporally segment the signal from the previous and subsequent noise dots, yielding adaptive temporal integration.
This PDF is available to Subscribers Only