Hansung Kim

University of Surrey, Kyungnam University

Papers

Total Citations

H-Index

About

Hansung Kim is a researcher at the forefront of computer vision and robotics, with key contributions spanning semantic scene understanding and real-time autonomous navigation. His most impactful work, "Semantic Scene Completion from a Single 360-Degree Image and Depth Map" (2020), introduces a novel deep convolutional neural network that leverages both synthetic and real RGB-D datasets to predict complete 3D semantic structures of indoor environments from a single panoramic view. This breakthrough has garnered 9 citations and addresses a critical limitation in prior scene completion methods, enabling more robust spatial reasoning for augmented reality and robotic perception. Earlier, Kim laid foundational groundwork in mobile robotics with "Real-time obstacle avoidance of mobile robots" (2007, 7 citations), where he developed a three-dimensional obstacle avoidance algorithm for spherical, omni-directional robots using point-based sensor data. This work demonstrated practical, real-time navigation solutions that remain relevant for autonomous systems. By bridging high-level semantic interpretation with low-level motion planning, Kim’s research empowers robots to both understand and safely traverse complex environments, marking him as a versatile innovator in intelligent systems.