This course explores the evolution of high-resolution mapping in field robotics over the past 25 years, focusing on applications in Marine Archaeology, Marine Geology, Fisheries Science, and Polar Studies. The course discusses the role of SLAM and other methodologies in mapping, highlighting their strengths and limitations. It also presents recent work on using light fields for robotics to address these limitations. The course covers topics such as optical imaging underwater, camouflage detection, machine learning for marine applications, emergency recovery, and dynamic objects in SLAM. The intended audience for this course includes researchers, students, and professionals interested in the intersection of robotics and marine exploration.
Mapping for Field Robotics - Hanumant Singh, Northeastern University
Paul G. Allen School via YouTube
Overview
Syllabus
SEMINAR SERIES WINTER 2021 INTERNET OF ROBOTIC THINGS
Marine and Polar Robotics - Where we have been, where we are and where we are going
Optical Imaging Underwater
The solution? SeaBed AUV
Camouflage - do you see the Skate?
Progress in breaking Camouflage
Segmentation
Classification
Vehicle Noise - Acoustic, Pressure Wave, Light
Machine Learning for Marine Applications
BIPOLAR ROBOTICS
Emergency Recovery
What did we Find?
The Second Law of Field Robotics
Low Contrast Environments
Melt Rate - Change Detection over Time
Dynamic Objects in SLAM-A Light Fields Approach
Acknowledgements
Taught by
Paul G. Allen School
