---

I study how minds learn the physical structure of the world from visual and sensorimotor experience. My research combines psychophysics, computational modeling, naturalistic egocentric video, and computer vision to ask how everyday interactions with objects shape perception, action, and intuitive physical reasoning.

A central goal of my work is to make “experience” measurable. Rather than treating prior experience as an abstract explanation for perception, I use large-scale naturalistic datasets to estimate the physical regularities people encounter as they look, move, and act in the world. I then test whether those measured regularities predict how people perceive motion, mass, force, and physical events.

🆕 New paper in Cognition

The Online Processing of Dynamics. Cognition, 270, 106416.
This paper develops the Online Processing of Dynamics model, which proposes that mass and velocity are represented jointly during dynamic event perception. Read the paper

🆕 New paper in Journal of Neurophysiology
Relational dynamics inform predictive motor planning and perception
Using a hybrid VR + real-object setup, we show that purely visual collision dynamics shape both how people plan their lifts and how heavy the objects feel. Relational motion cues create a dynamic weight illusion: objects that “should” be heavier based on the collision feel heavier, even when their physical mass is identical.
Read the paper