Neural differences between chromatic- and luminance-driven attentional salience in visual search

Abstract

Previous electroencephalographic research on attentional salience did not fully capture the complexities of low-level vision, which relies on both cone-opponent chromatic and cone-additive luminance mechanisms. We systematically varied color and luminance contrast using a visual search task for a higher contrast target to assess the degree to which the salience-computing attentional mechanisms are constrained by low-level visual inputs. In our first experiment, stimuli were defined by contrast that isolated chromatic or luminance mechanisms. In our second experiment, targets were defined by contrasts that isolated or combined achromatic and chromatic mechanisms. In both experiments, event-related potential waveforms contralateral and ipsilateral to the target were qualitatively different for chromaticcompared to luminance-defined stimuli. The same was true of the difference waves computed from these waveforms, with isoluminant stimuli eliciting a mid-latency posterior contralateral negativity (PCN) component and achromatic stimuli eliciting a complex of multiple components, including an early posterior contralateral positivity followed by a late-latency PCN. Combining color with luminance resulted in waveform and difference wave patterns equivalent to those of achromatic stimuli. When large levels of chromaticity contrast were added to targets with small levels of luminance contrast, PCN latency was speeded. In conclusion, the mechanisms underlying attentional salience are constrained by the low-level inputs they receive. Furthermore, speeded PCN latencies for stimuli that combine color and luminance signals compared to stimuli that contain luminance alone demonstrate that color and luminance channels are integrated during pre-attentive visual processing, before top-down allocation of attention is triggered.

Color needs luminance for visual selection during scene search

“…The most parsimonious interpretation in the context of earlier findings is that peripheral color is beneficial for search, but only if it is supported by luminance. This interpretation is in accord with Hardman et al (2020), who found in a task with fixation demands that luminance contrast added to chromaticity contrasts speeded up an event-related EEG component, in comparison to pure luminance contrasts. Color and luminance might be integrated during pre-attentive processing, but luminance dominates.…”

Salience of spatiochromatic patterns

“…Finally, bluish required less physical contrast to achieve equal perceived contrast with the standard compared to yellowish only for Gaussians. These Gaussian-specific effects replicate the findings of Hardman et al (2020) , which observed the same effect using contrast matching with a stimulus that consisted of eight small Gaussians. The differences between bluish and yellowish Gaussians persisted even after adjustment using DTs.…”

“…The unique spatial properties of the Gaussian cause it to have a generalized difference in contrast matches compared to gratings and Gabors and a specific asymmetry in perceived contrast between bluish and yellowish (bluish requiring less perceived contrast than yellowish to achieve equal perceived contrast with a standard, as observed earlier by Hardman et al, 2020 ). Gaussians more strongly activate single-opponent neurons and lowest spatial frequency channels while gratings and Gabors more strongly activate higher spatial frequency channels and double-opponent neurons.…”

“…In particular, we aimed to establish whether bluish would still require more physical contrast to achieve equal perceived contrast with yellowish for patterns other than gratings (as in Switkes, 2008 ). This curiosity was at least in part driven by our observation of the opposite result (bluish more salient than yellowish) in a study in which we used salience matching to equate the appearance of small bluish or yellowish Gaussian patches to a light gray standard ( Hardman, Töllner, & Martinovic, 2020 ).…”

Luminance Contrast Drives Interactions between Perception and Working Memory

“…discriminate circularity of a shape; Mullen & Beaudot, 2002) on stimuli defined by luminance, reddish-greenish or bluish-yellowish modulations. The resulting contrasts are then expressed in units of detection threshold for each mechanism, to control for more basic differences in threshold sensitivity, which also has relevance for suprathreshold salience (Kulikowski, 1976; for more recent work, see Hardman, Toellner &Martinovic, 2020 andMartinovic, 2021). We adopted this approach in the first experiment, a psychophysical delayed recognition threshold task.…”