Greg Izatt

Vassar College, Massachusetts Institute of Technology

Papers

Total Citations

H-Index

About

Greg Izatt is a leading researcher in human-robot interaction and shared autonomy, with a focus on designing intuitive interfaces for complex robotic systems. His most influential work, "Director: A User Interface Designed for Robot Operation with Shared Autonomy" (2016, 49 citations), introduces a novel interface that enables operators to manage high-degree-of-freedom mobile manipulators—such as humanoids—in field scenarios. This system enhances situational awareness and facilitates interactive motion planning and control, allowing operators to delegate tasks while retaining supervisory oversight. Izatt’s contributions address critical challenges in teleoperation, reducing operator burden and improving robot efficiency in real-world applications. In his 2019 paper "A Supervised Approach to Predicting Noise in Depth Images" (29 citations), Izatt tackles the problem of sensor noise in depth imagery, a key issue for robotic simulation and testing. By developing a supervised learning method to predict pixel-wise dropouts and distortions, he enables more accurate simulation environments for verifying robotic behavior. His work has significant implications for the reliability of autonomous systems in complex settings. With over 87 total citations, Izatt’s research bridges interface design and perception, advancing the practical deployment of robots in unstructured environments. His achievements underscore a commitment to making robots more accessible and effective through thoughtful human-centered design.