Matt Deitke

Hi! I am interested in computer vision, reinforcement learning, and embodied AI. I work at the Allen Institute for AI and am an undergraduate at the University of Washington in the Paul G. Allen School of Computer Science.

Previously, I spent my early teenage years as a computer graphics developer in the Department of Athletics at The Ohio State University, the University of Cincinnati, and several startups. Later in high school, I studied machine learning and deep learning at Georgia Tech.


RoboTHOR is an environment within the AI2-THOR framework, designed to develop embodied AI agents. It consists of a series of scenes in simulation with counterparts in the physical world.

Learn More

AI2-THOR is a highly customizable interactive physics simulation platform used to facilitate research in computer vision and embodied AI.



Embodied AI

CVPR 2020

I co-hosted the Embodied AI workshop at CVPR 2020. The goal of the workshop was to bring together researchers from the fields of computer vision, language, graphics, and robotics to share and discuss the current state of intelligent agents that can see, talk, act, and reason.





Papers & Books: Omitted.


These are some notes I took while getting started in computer vision and machine learning while in high school. I never officially took any of these courses (except the well-taught grad version of UW's CSE 455 when I got to UW):

CVPR Paper Viewers

Mainly build and maintained for myself, to quickly skim topics in papers at CVPR [code]: