Ensemble coding of color and luminance contrast

  • PDF / 1,106,691 Bytes
  • 14 Pages / 595.276 x 790.866 pts Page_size
  • 84 Downloads / 208 Views

DOWNLOAD

REPORT


Ensemble coding of color and luminance contrast Siddhart Rajendran 1,2 & John Maule 3 & Anna Franklin 3 & Michael A. Webster 1 Accepted: 30 August 2020 # The Psychonomic Society, Inc. 2020

Abstract Ensemble coding has been demonstrated for many attributes including color, but the metrics on which this coding is based remain uncertain. We examined ensemble percepts for stimulus sets that varied in chromatic contrast between complementary hues, or that varied in luminance contrast between increments and decrements, in both cases focusing on the ensemble percepts for the neutral gray stimulus defining the category boundary. Each ensemble was composed of 16 circles with four contrast levels. Observers saw the display for 0.5 s and then judged whether a target contrast was a member of the set. False alarms were high for intermediate contrasts (within the range of the ensemble) and fell for higher or lower values. However, for ensembles with complementary hues, gray was less likely to be reported as a member, even when it represented the mean chromaticity of the set. When the settings were repeated for luminance contrast, false alarms for gray were higher and fell off more gradually for out-ofrange contrasts. This difference implies that opposite luminance polarities represent a more continuous perceptual dimension than opponent-color variations, and that “gray” is a stronger category boundary for chromatic than luminance contrasts. For color, our results suggest that ensemble percepts reflect pooling within rather than between large hue differences, perhaps because the visual system represents hue differences more like qualitatively different categories than like quantitative differences within an underlying color “space.” The differences for luminance and color suggest more generally that ensemble coding for different visual attributes might depend on different processes that in turn depend on the format of the visual representation. Keywords Color and light . Color . Color and light . Contrast, color and light . Lightness/brightness

Introduction In complex images, observers are often more sensitive to the gist of the scene than to the individual items composing the scene (Alvarez, 2011; Ariely, 2001; Whitney & Yamanashi Leib, 2018). These summary percepts are called ensemble coding, and have been demonstrated for a number of stimulus features including features like size (Ariely, 2001), motion (Watamaniuk & Duchon, 1992), and orientation (Parkes, Lund, Angelucci, Solomon, & Morgan, 2001), as well as high-level attributes like faces (Elias, Dyer, & Sweeny, 2017; Haberman, Harp, & Whitney, 2009; Haberman & Whitney, 2007), biological motion (Sweeny, Haroz, & Whitney, 2013), or “lifelikeness” (Leib, Kosovicheva, & Whitney, 2016). In each of these cases observers can readily

* Michael A. Webster [email protected] 1

Department of Psychology, University of Nevada, Reno, NV, USA

2

LV Prasad Eye Institute, Hyderabad, India

3

School of Psychology, University of Sussex, Falmer, Sussex, UK

extract the average of the stimu