Sid Bao earns Best Student Paper Award for Computer Vision Research

Bao’s research is in Semantic Structure from Motion, a new framework for jointly recognizing objects as well as reconstructing their underlying 3D geometry.

sid bao Enlarge
Sid Bao at his computer, which displays a scene showing his algorithm being applied to a specific scene. Using this algorithm, a robot would be able to identify and localize buildings, roads, the sky, or other objects.

Sid Ying-Ze Bao, an electrical engineering:systems PhD student, was first author on a winning paper at the 1st IEEE Workshop on Challenges and Opportunities in Robot Perception, held in Barcelona, Spain in conjunction with the International Conference on Computer Vision. He received the Best Student Paper Award for the paper, Semantic Structure From Motion with Object and Point Interactions, co-authored by EE:Systems master’s student Mohit Bagra and Bao’s advisor, Prof. Silvio Savarese.

Mr. Bao describes the research:
Semantic Structure from Motion (SSFM) is a new framework for jointly recognizing objects as well as reconstructing the underlying 3D geometry of the scene from multiple semi calibrated images. In SSFM we explore the interaction among scene elements across different viewpoints in order to: i) estimate camera poses from images; ii) accurately localize objects in 3D; and iii) significantly improve detection and reconstruction results over existing state-of-the-art methods.
Mr. Bao’s research interests include: 1) image-based geometry reconstruction with high level understanding; 2) real time image understanding and object detection; and 3) mobile application of computer vision.

The reseach was funded in part by Prof. Savarese’s NSF CAREER award, and the Gigascale Systems Research Center.