Abstract
A fundamental question in the visual neurosciences relates to how high-level visual patterns are extracted from a visual scene and segregated into discrete object categories. Extensive literature has been dedicated to the investigation of the cognitive and neural processes through which object categories are formed, but little is known of how complex visual object categories are extracted from a crowded visual environment and then represented in the cortical visual hierarchy. Here, we used event-related potentials (ERPs) to investigate the neural underpinnings of visual object category extraction in a cluttered visual field. EEG was continuously recorded while observers were given a hybrid of an object category learning and visual search task. In this hybrid task, a peripheral array of four dot patterns was briefly flashed for 200 ms. In 50% of the trials, one position of the peripheral array contained a distortion of a prototype dot pattern. The visual array in the remaining trials consisted entirely of randomly generated dot patterns. Over the course of hundreds of trials, observers learned to detect the dot pattern object category via correct or incorrect feedback given on each trial. We assessed these improvements in dot pattern detection (d’) in conjunction with visual ERPs to examine the neural mechanisms of visual object category formation. Preliminary results indicate that observers exhibited significant improvements in dot pattern detection over the course of the experiment and that these improvements were associated with the emergence of a visually evoked component previously linked to category learning and perceptual expertise: the N250. These results elucidate the neural mechanisms underlying the formation of cortical representations for novel visual object categories and the processes through which these object categories are extracted from a complex visual array.
Meeting abstract presented at VSS 2015