Title: Human Pose Estimation in Bed

Date: Tuesday, June 2nd, 2020

Time: 1:00 PM - 2:30 PM (EST)

Location: BlueJeans meeting (https://bluejeans.com/386994880/6157)

Henry M. Clever

Robotics Ph.D. Student

Department of Biomedical Engineering

Georgia Institute of Technology

Committee:

Dr. Charlie Kemp (Advisor) – Department of Biomedical Engineering, Georgia Institute of Technology

Dr. James Hays – College of Computing, Georgia Institute of Technology

Dr. Ayanna Howard – Department of Interactive Computing, Georgia Institute of Technology

Dr. C. Karen Liu – Department of Computer Science, Stanford University

Dr. Greg Turk – Department of Interactive Computing, Georgia Institute of Technology

Abstract:

People spend a substantial part of their lives at rest in bed. 3D human pose and shape estimation for this activity would be beneficial to numerous applications, including remote patient care, bed sore management, and assistive robotics. However, this is a challenging perception problem due to a variety of factors, including bedding covering the body, nearby medical equipment, and the unavailability of well-labeled perceptual data. To overcome these challenges, we use a pressure sensing array on the bed to sense the body in a manner that is insensitive to bedding, and physics simulations to automatically generate synthetic perceptual data at scale with labels. We also develop novel deep learning models, including a model that infers body shape and pose from a real pressure image when trained exclusively on synthetic data. For the remainder of this dissertation, we propose new investigations into the use of a depth sensing camera above the bed to complement the pressure sensing array, and the use of our estimation methods for assistive robotics application.