Small Study Investigates Automatic Object Discrimination without Visual Attention

Article: Experiment Shows How Our Visual System Avoids Overloading
Source: HSE University (Russia), via Technology Networks
Published: February 9, 2021

Studies into vision science are notoriously difficult to extrapolate data from due to the complexity of neural processing. In this study, junior researchers at HSE University in Russia sought to determine whether our visual system parses spatially mixed objects of different categories automatically or whether such parsing requires attention. For example, when we look at an apple tree, we easily differentiate the apples from the leaves, but how much attention is involved? Intuitively, it makes sense that much of visual processing would be automatic, given the wealth of visual input and the bottlenecks of conscious attention. The present experiment investigated this hypothesis by loading participants' attention with a central task (such as looking at a cross) while unattended background visual stimuli (such as lines of various lengths and orientations) changed. Markers of automatic sensory discrimination were measured by visual mismatch negativity (vMMN) via electroencephalography (EEG). What can be extrapolated from vMMN results, such as multiple peaks indicating category separation, is beyond the scope of this review. A major limitation of the study, however, is its exceedingly small sample size of only 20 participants; nonetheless, the conclusion that object discrimination can occur without attention is not surprising.

Electrophysiological results of the experiments

Personal commentary: I would add that at least in experiments in monkeys, while attention is not required to evoke a neuronal response, attending to the stimulus evokes a greater neuronal response. And when a stimulus is both involuntary and voluntary, such as the abrupt onset of a visual stimulus in a neuron's receptive field or the planning of a saccade eye movement to bring the stimulus into the neuron's receptive field, then an even stronger neuronal response is evoked. Furthermore, these neuronal responses are enhanced when the stimulus is relevant to the animal's behavior.

My rating of this study: ⭐

Khvostov VA, Lukashevich AO and Utochkin IS. "Spatially intermixed objects of different categories are parsed automatically." Scientific Reports.  11(377). 11 January 2021. https://doi.org/10.1038/s41598-020-79828-4