Purchase this article with an account.
Chien-Chung Chen, Charles Chubb; Anchoring of “black” in texture discrimination. Journal of Vision 2015;15(12):772. doi: https://doi.org/10.1167/15.12.772.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The visual system is biased toward negative luminance contrast. In particular, texture discrimination performance is dominated by a “blackshot” mechanism that is sensitive to the lowest luminances in the image (Chubb et al. 2004, Vision Research). However, it is unclear whether the tuning of this mechanism is absolute (driven by luminance near black) or adaptive (driven by the lowest luminance in the current input). To investigate this issue, we used a task in which the participant strove to classify white noise textures according to whether the histogram on a given trial had higher or lower variance than the uniform histogram. There were five conditions, each of which used 9 luminances: (1) full gamut: stimuli included luminances from black to white; (2) high gamut: stimuli included luminances from dark gray to white; (3) low gamut: stimuli included luminances from black to light gray; (4) uninformative dark: stimuli included luminances from black to white with the proportions of the lowest two luminances fixed at 1/9 on each trial; and (5) uninformative bright: stimuli included luminances from black to white with the proportions of the highest two luminances fixed at 1/9 on each trial. The threshold histogram differences for the full, low and high gamut textures were similar, and the estimated sensitivity functions, scaled by luminance range, were identical, despite the fact that the mean luminance of the high gamut double that of the low gamut. The observer had difficulty in discriminating uninformative dark pattern with a 2-fold increase in modulation threshold. Our results suggest that the tuning of the blackshot mechanism is adaptive, not absolute: the observers’ judgments are anchored to the lowest luminance in the image, not to black.
Meeting abstract presented at VSS 2015
This PDF is available to Subscribers Only