Ph.D Project- Jason Armitage

Recent work in machine learning opens the way to perform analysis on changes in learned spatial structures over time with embodied AI, that is by placing artificial agents in virtual environments. Consider an agent that must navigate a virtual rendering of a location using instructions, visual cues, and spatial information derived from a predefined set of sources. In the performance of actions based on linguistic and visual inputs, the agent interacts with the environment to develop and display spatial abilities that are shaped by the location and materials that it learns from. The proposed research will develop methods and tools to support empirical investigation of changes in representation and environments over time and location.