Xueying Ren¹ (), Kendrick Kay², James Booth¹, Franco Pestilli³, Sophia Vinci-Booher¹; ¹Dept. of Psychology and Human Development, Vanderbilt University, Nashville, Tennessee, USA, ²Dept. of Radiology, University of Minnesota, Minneapolis, Minnesota, USA, ³Dept. of Psychology, University of Texas at Austin, Austin, Texas, USA

Cortical regions within the ventral visual stream demonstrate category selectivity, with areas such as the fusiform face area showing increased activity when viewing faces compared to other categories. Traditional methods for localizing category-selective responses typically use isolated stimuli on a gray background with a central fixation, potentially limiting our understanding of how categories are processed during real-world visual experiences. This study explored how the human brain processes four different visual categories embedded in complex scenes under free-viewing conditions to offer insight into how the brain processes and integrates visual information in real-world settings. To this end, we collected fMRI data from 10 adult participants as they viewed images selected from the Sesame Street Archive, which includes a prevalence of categories uncommon in other image datasets. Each image featured at least one of four categories: faces, places, words, and numbers, which are prevalent in educational and everyday contexts. Univariate analyses at the subject level revealed individual variability in brain activations for different categories. Notably, faces and places elicited relatively more consistent activation patterns across participants, whereas representations for words and numbers exhibited larger individual differences. These findings suggest that certain category-selective brain responses, particularly for faces and places, can be effectively captured using complex scenes under free-viewing conditions. However, we observed individual variability in responses to learned object categories, such as words or numbers, which may reflect differences in reading and math abilities. By using naturalistic stimulation with rich visual contexts, this research provides new insights into the nuances of visual processing at the individual level.

Acknowledgements: 5R01HD114489