Minsu Kim

Minsu Kim

I am a second-year Ph.D. student in the School of Electrical Engineering at KAIST, advised by Prof. Junmo Kim. Before beginning my doctoral studies, I worked as a specialist at KIST and was a research intern at NAVER LABS, where I was mentored by Prof. Giseop Kim and Dr. Sunwook Choi. I received my M.S. from DGIST under the supervision of Prof. Kyong Hwan Jin and completed my undergraduate studies at Pusan National University.

My research agenda focuses on making foundation models fast, memory-efficient, and affordable across both training and inference. I approach this goal from a statistical and systems perspective, investigating low-bit and memory-efficient ML systems for scalable training and deployment.

Email | LinkedIn | Google Scholar | GitHub

Publications

InlierQ

Inlier-Centric Post-Training Quantization for Object Detection Models

Minsu Kim, Dongyeun Lee, Jaemyung Yu, Jiwan Hur, Giseop Kim, and Junmo Kim

International Conference on Learning Representations (ICLR), 2026

[Paper]

BEVRestore

Addressing Diverging Training Costs Using BEVRestore for High-Resolution BEV Map Construction

Minsu Kim, Giseop Kim, and Sunwook Choi

IEEE Robotics and Automation Letters (RA-L), 2024

[Paper] [Code] [Project]

BroadBEV

BroadBEV: Collaborative LiDAR-camera Fusion for Broad-sighted BEV Map Construction

Minsu Kim, Giseop Kim, Kyong Hwan Jin, and Sunwook Choi

IEEE International Conference on Robotics and Automation (ICRA), 2024

[Paper] [Code]

Image Stitching

Learning Residual Elastic Warps for Image Stitching under Dirichlet Boundary Condition

Minsu Kim, Yongjun Lee, Woo Kyoung Han, and Kyong Hwan Jin

IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2024

[Paper] [Code]

Image Stitching

Implicit Neural Image Stitching With Enhanced and Blended Feature Reconstruction

Minsu Kim, Jaewon Lee, Byeonghun Lee, Sunghoon Im, and Kyong Hwan Jin

IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2024

[Paper] [Code]

Awards and Honors