Research

The SCCN lab studies how humans recognize other people and acquire knowledge about them, as well as disorders of social cognition, using a combination of behavioral studies, computational models, and neuroimaging. Research is organized along the following main focus areas:

Social perception

In previous work, we have identified neural representations that encode information about the identity of a face, and about the identity of a person across different perceptual modalities. With Emily Schwartz, Hamed Karimi, Mo Zhou, Jianxin Wang, Christina Farmer and Nicholas Arangio (in collaboration with Avniel Ghuman, Arish Alreja and Mark Richardson, and with the support of a NSF CAREER grant)  we are currently using deep networks to study the computational mechanisms for the recognition of face identity and facial expressions, and for the recognition of actions and bodies, from static and dynamic stimuli.

Person knowledge

We are interested in the cognitive and neural mechanisms with which knowledge about others is acquired and updated. With Hamed Karimi, Gabriel Fajardo, and Obinna Onyekachiuzoamaka we are studying how observers understand and predict the behavior of others, using inverse reinforcement learning (IRL) models and Bayesian Nonparametric models. With Minjae Kim and Tony Chen (in collaboration with Liane Young), we are investigating how we represent the traits of other famous people and of strangers for whom we only know a single behavior, and we are studying how representations of traits are updated taking into account the situations.

Autism spectrum disorder

With Aidas Aglinskas, Alicia Bergeron, Julianna Pijar, and Yu Zhu (in collaboration with Joshua Hartshorne, and with the support of the Simons Foundation Autism Research Initiative) we are studying individual differences within Autism Spectrum Disorders (ASD) using deep learning methods to disentangle individual variation specific to ASD from variation shared with the general population.

Multivariate and nonlinear interactions between brain regions

With Mengting Fang, we have developed a toolbox for multivariate pattern dependence (PyMVPD, in press). With Craig Poskanzer we have tested nonlinear models of the interactions between brain regions. With Mengting Fang and Gabriel Fajardo, we are using these methods to study the co-occurrence of information from multiple sensory modalities and about objects from multiple categories within common brain regions.