FriSem

Beth Rispoli, Ph.D. Student in Psychology, admitted Autumn 2022

Title: Dynamics and Category-Selectivity in High-Level Visual Processing

Abstract: Efficient visual perception is essential for navigating and interacting with the world around us. The brain has evolved to include optimized processing streams for important visual functions. The ventral stream processes vision for recognition, while the dorsal stream is involved with spatial processing. Recently, evidence has been found supporting the existence of an anatomically and functionally distinct third visual processing stream along the lateral brain surface. Projecting from early visual areas to the lateral occipitotemporal cortex (LOTC) and superior temporal sulcus (STS), this pathway has been proposed to process various aspects of motion, action perception, dynamic social cues, and multimodal information. While these proposals provide insight into overarching functional goals, little has been done to systematically characterize the intricacies of lateral stream processing and how it can be differentiated from similar ventral stream processing. Similar to the ventral stream, the lateral stream has been found to include topographically organized category-specific areas that show preferences for face, limb, object, and scene stimuli. Differences in the processing of stimuli that change rapidly over time may be a factor in the ventral-lateral distinction. Available stimuli and neuroimaging datasets have been a limiting factor in furthering our understanding of these ventral-lateral stream distinctions. The solution to this problem is to create a novel stimuli set and large-scale dataset that has the ability to disentangle differences in processing between ventral and lateral regions for a range of dynamic stimuli that can capture variations in motion, movements, complexities of contextual information, and category. In this FYP talk, I will discuss the design and creation of a novel stimuli set that contains over 4,800 individual video stimuli, the questions we intend to answer using it, and present pilot data aimed at validating the experiment’s usefulness.