The Importance of Deep Learning in Robotics Research

Abstract

The popularity of deep learning among roboticists is not evidence that it is a fruitful path to a greater scientific understanding into open problems in robotics. Nonetheless, I will argue that the pervasiveness of deep learning in robotics research is not an impediment to scientific insights into robotics problems, but rather leads to a probabilistic and data-driven perspective that will be essential for solving the next generation of robotics challenges.

Date
May 24, 2019 11:15 AM — 11:25 AM
Location
Montréal, Canada
Stewart Jamieson
Stewart Jamieson
PhD Candidate

My research interests include semantic mapping and path planning for multi-agent underwater robotic systems.