Seho KIM (김세호)

Education

• 2021.3 - Present        Ph.D. (School of Electrical Engineering), Korea Advanced Institute of Science Technology, South Korea

• 2019.3 - 2021.2          M.S. (School of Electrical Engineering), Korea Advanced Institute of Science Technology, South Korea

• 2014.3 - 2019.2          B.S. (School of Electrical Engineering), Korea Advanced Institute of Science Technology, South Korea

Research Interests

• Computer Vision

• Deep Learning

• Visual Odometry

Projects

Publications

• Ue-Hwan Kim, Se-Ho Kim and Jong-Hwan Kim, "SimVODIS: Simultaneous Visual Odometry, Object Detection, and Instance Segmentation," in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 1, pp. 428-441, 1 Jan. 2022, doi: 10.1109/TPAMI.2020.3007546.

• Ue-Hwan Kim*, Se-Ho Kim* and Jong-Hwan Kim, "SimVODIS++: Neural Semantic Visual Odometry in Dynamic Environments," in IEEE Robotics and Automation Letters, vol. 7, no. 2, pp. 4244-4251, April 2022, doi: 10.1109/LRA.2022.3150854. (* These two authors contributed equally)

• Inyong Koo*, Inyoung Lee*, Se-Ho Kim, Hee-Seon Kim, Woo-jin Jeon, and Changick Kim. “PG-RCNN: Semantic Surface Point Generation for 3D Object Detection,” Accepted to IEEE/CVF International Conference on Computer Vision (ICCV), 2023. (* These two authors contributed equally)

• Se-Ho Kim*, Inyong Koo*, Inyoung Lee, Byeongjun Park, and Changick Kim. "DiffRef3D: A Diffusion-based Proposal Refinement Framework for 3D Object Detection," Submitted to AAAI, 2024 (* These two authors contributed equally)