Tag
This paper presents a model-guided framework using AI to discover and generate facial emotion stimuli that maximize perceptual differences between autistic and neurotypical individuals, demonstrating that group differences are concentrated in a small subset of expressions.
GEESE is an end-to-end deep learning framework that learns behavioral representations directly from 3D pose dynamics without hand-crafted features, surpassing traditional baselines in behavior classification and genotype prediction across three autism-associated genetic models. It also introduces HONK, an interactive tool for natural language-based behavioral phenotyping.