projects
project thumbnail

Modeling Dynamic Social Vision Highlights Gaps Between Deep Learning and Humans

We present a dataset of natural videos and captions involving complex multi-agent interactions, and we benchmark 350+ image, video, and language models on behavioral and neural responses to the videos. Together these results identify a major gap in AI's ability to match the human brain and behavior and highlight the importance of studying vision in dynamic, natural contexts.
project thumbnail

Large-scale Deep Neural Network Benchmarking for Dynamic Social Vision

This study systematically benchmarks over 300 deep neural networks (DNNs) against fMRI responses to naturalistic social videos to evaluate their correspondence with the brain's lateral visual stream, a region implicated in dynamic social vision.
project thumbnail

Predicting Dimensional Symptoms of Psychopathology from Task-Based fMRI using Support Vector Regression

This study develops a novel machine learning approach using Support Vector Regression (SVR) to explore potential biomarkers in fMRI data for symptoms of anxiety and depression, finding that MID task-fMRI data does not accurately predict these symptoms, with results indicating a poor model fit.