Biography | Publications

Dr. Oh received his Ph.D. (2009) and M.S.(2008) in Computer Science from Georgia Institute of Technology and his B.S. in CS from Seoul National University (2003). He is a member of the Computer Vision Group at Kitware, Inc.

Dr. Oh has worked on a wide array of technical problems which include computer vision, robotics, AI, pattern recognition, machine learning, augmented reality, data mining and wearable computing. He is interested in broad research topics related to building intelligent systems where traditional AI meets modern data-driven approaches. He has published more than 15 papers in academic venues such as IJCV, CVPR, ICCV, ECCV, ICPR, ICRA, and IROS, and has reviewed papers for journals such as IEEE T-PAMI, IEEE T-SP and CVIU. His latest work with Kitware has focused on developing technical approaches to understand/recognize events and activities from challenging video collections - he has been a technical contributor to programs such as DARPA 'Video and Image Retrieval and Analysis Toolkit (VIRAT)', DARPA 'Persistent Stare Exploitation and Analysis System (PerSEAS)', and numerous SBIRs/STTRs such as the DARPA SBIR 'Functional Object Recognition', DARPA STTR 'Building Labeling in Urban Area (BLUE)', and ONR SBIR 'Human Activity Discovery in Video'.

Previously, he developed various extensions of switching linear dynamic systems, which have been successfully applied to tracking and activity understanding problems for both computer vision and wearable computing, for which he was awarded a prestigious Samsung research fellowship (2003-2007). Additionally, he was a primary contributor for the GPS-based localization algorithm used in the System for Wearable Audio Navigation (SWAN) developed at Georgia Tech (2004); he co-developed the first affordance-based robotic vision algorithms for outdoor robot navigation as part of the DARPA 'Learning Applied to Ground Robotics (LAGR)' program (2005-2007), and developed a fast facial landmark detection method with his collaborators at Microsoft Research (U.S. patent, 2006). Additionally, he has been involved in diverse research programs which involve robotic visual localization and mapping, dynamic augmented earth visualization project (featured in mainstream media such as CNN, Wired etc, 2008), and video retrieval projects for child healthcare solutions (2009).

Publications