Kyumin Hwang

Hi, I'm Kyumin Hwang, a Ph.D. Student in the Department of Electrical Engineering and Computer Science at DGIST, South Korea. I am conducting my research under the supervision of Prof. Sunghoon Im at the DGIST Computer Vision Lab.

Previously, I earned the B.S. degree from the Department of Computer Science and Engineering, Kyungpook National University, in 2020, and the M.S. degree from the Department of Information and Communication Engineering, DGIST, in 2022.

My research interests include 3D computer vision, particularly multi-view stereo for 3D reconstruction, and deep learning with geometry for autonomous driving. Recently, I have also been focusing on 3D scene understanding, including occupancy prediction and rendering.

Email  /  Google Scholar  /  Github  /  Curriculum Vitae

2025

* indicates equal contribution
† indicates corresponding author

We propose a reflective-aware triplet loss and multi-teacher knowledge distillation, improving depth accuracy on reflective surfaces.

We proposed a training strategy via intrinsic image decomposition to reduce errors on reflective surfaces and improve the depth prediction accuracy.

We investigate the generalization performance of monocular depth estimation using various backbone networks, including CNN and Transformer models, by analyzing texture and shape biases through extensive experiments on both synthetic and real-world datasets.

Website Template