Contributed Talk Sessions | Poster Sessions | All Posters | Search Papers

Poster Session A: Tuesday, August 12, 1:30 – 4:30 pm, de Brug & E‑Hall

Feature-based Modeling of Visual Attention in Autism: A Large-scale Online Eye-tracking Study

Qianying Wu¹, Na Yeon Kim¹, Audrey K. Lai, Lynn K. Paul¹, Ralph Adolphs¹; ¹California Institute of Technology

Presenter: Qianying Wu

Atypical visual attention is one of the most reliable findings in autism spectrum disorder (ASD), with important implications for clinical screening and diagnosis. However, most findings rely on artificial stimuli and small samples, limiting generalizability. In this pre-registered study, we used webcam-based eye-tracking and feature-based computational modeling to characterize visual attention in a broad sample of 336 ASD participants and 304 neurotypical controls. Participants watched videos of group conversations that incorporated controlled social and nonsocial features. Compared to controls, autistic individuals showed reduced attention to speakers, increased sensitivity to distractors, and more frequent gaze shifts. Our study demonstrates the power of scalable online eye-tracking and modeling approaches for capturing individual differences in visual attention and advancing the understanding of heterogeneity in ASD.

Topic Area: Object Recognition & Visual Attention

Extended Abstract: Full Text PDF