AI-guided stimuli discovery and generation to optimize facial emotion perception studies in autism
2026-07-09 • Artificial Intelligence
Artificial IntelligenceMachine Learning
AI summaryⓘ
The authors studied how autistic and non-autistic adults perceive facial emotions differently. They found that these differences appear strongly for certain specific faces rather than evenly across all images. By using special computer models, they identified and created faces that highlight these differences or make perceptions more similar between groups. This approach helps scientists better understand and test how people with different brains see emotions on faces. Overall, the authors suggest a way to improve tests by focusing on the most informative images instead of averaging responses on many pictures.
autismfacial emotion perceptionneurotypicalbehavioral assaysartificial neural networksgenerative adversarial networkphenotypestimulus optimizationpopulation differences
Authors
Kushin Mukherjee, Na Yeon Kim, Maren Wehrheim, Ralph Adolphs, Kohitij Kar
Abstract
Understanding perceptual differences between autistic and neurotypical adults requires behavioral assays that are sensitive, reliable, and mechanistically informative. Facial emotion perception is a useful test case because group differences have been reported, but findings vary across studies. Here we show that this variability may reflect image-level sparsity: autistic-neurotypical differences in emotion judgments were concentrated in a small subset of diagnostic facial expressions rather than spread uniformly across stimuli. We trained population-specific artificial neural network models to predict image-level judgments for autistic and neurotypical participants, then used these models to select novel faces predicted to maximize group separation. In an independent cohort, model-selected images produced larger behavioral differences than matched random images. We then used the same models with a generative adversarial network to transform diagnostic images toward greater predicted group agreement. In phenotype-matched validation, synthesized images reduced behavioral separation relative to their matched originals. These results establish a model-guided framework for discovering and transforming stimuli that reveal population-specific perceptual differences. More broadly, they show how behavioral phenotyping can move beyond averaging across fixed stimulus sets toward optimized assays that identify the conditions under which neurodivergent perception diverges or converges.