SJ Kim

I am an M.S. candidate at Kyung Hee University's Visual Science Lab (VSLab), advised by Prof. MyeongAh Cho.

My research centers on video understanding and 3D perception for intelligent systems. I focus on multi-object tracking with adaptive motion prediction, multi-modal 3D object detection through sensor fusion, and novel view synthesis using Gaussian Splatting. I am also exploring few-shot object detection to enable efficient learning in data-scarce scenarios.

My work aims to advance robust and generalizable computer vision technologies for understanding dynamic 3D environments in real-world applications.

Email  /  CV  /  Github  /  Google Scholar

News

[Feb. 2026] Received the 🏆 Best Presentation Award at KSC 2025!

[Nov. 2025] One paper got accepted at AAAI 2026! (Acceptance rate: 17.6%)

Research

PlugTrack is a plug-and-play motion prediction framework that adaptively fuses Kalman filter and any data-driven motion predictors for robust multi-object tracking.

This website's source code is borrowed from Jon Barron's website template.