Understanding objects in the real world requires recognizing them even as our sensory input constantly changes. Most research—both in biology and AI—has focused on how objects are represented using static information, but real environments are dynamic. This project aims to uncover how robust object representations form and evolve under changing conditions. The team will (1) develop new quantitative methods to study object representations in dynamic settings, (2) apply these tools to investigate how biological systems handle these changes, and (3) design and test new dynamic neural network models that better mirror real-world object perception.