Jong Woo Im
Seoul National University · Computer Science
About the Lab
Professor Jong Woo Im's research lab specializes in advanced materials for energy storage and cutting-edge computer vision techniques for 3D reconstruction and robotics. The lab explores high-nickel, cobalt-free cathode materials for next-generation lithium-ion batteries, focusing on mitigating chemo-mechanical degradation through multi-scale characterization. In parallel, the lab develops innovative algorithms for non-rigid 3D object alignment from RGB-D data and robust video stabilization using LIDAR-inertial odometry, particularly for autonomous navigation in challenging environments. A key focus is also on wide-baseline, multi-camera systems with ultra-wide-angle lenses, enabling high-precision calibration and 3D sensing for robotics and 3D modeling applications.
Research Overview
Research Output Trend
Figures are computed from collected data and may differ slightly.
Selected Papers
12The present paper attempts to construct a concise history of electroacoustic music in Korea. It will be shown that this history is best represented as the conflicting ideals of what constitutes electroacoustic music. Drawing on the interviews with key figures from the past and the present, active in both academic and ‘underground’ scenes, the socio-political nature of the history is excavated. This discourse, in the Foucauldian sense, makes it possible to contextualise electroacoustic music in K
Recording of Jongwoo Yim's Absorption. This work includes instrumental sounds as well as electronic sounds (both from tape and live electronics). The clarinet line is a spotlight and eventually gets pulled into the harmonies of the other parts.
The computer music studios of the composition department at Hanyang University in
Recording of Jongwoo Yim's Respiration I et II. This work is compromised of the first two etudes in a collection of four that are designed for percussion and electronics. The first etude is dedicated to percussion, especially skin instruments, and electronics. The second etude is composed of voice and percussion instruments.
High-Ni layered oxide cathodes without Co are being investigated as potential cathode materials for Li- ion batteries with high energy density. By decreasing the Co content, these cathodes not only boost energy density but also alleviate concerns about the supply instability and fluctuating cost of Co raw materials. However, the elevated Ni content in the layered oxides causes distinct chemo-mechanical degradation mechanisms that inhibit their commercial application. In order to gain insight int
Background Wound healing is an interaction of a complex signaling cascade of cellular events, including inflammation, proliferation, and maturation. K+ channels modulate the mitogen-activated protein kinase (MAPK) signaling pathway. Here, we investigated whether K+ channel-activated MAPK signaling directs collagen synthesis and angiogenesis in wound healing. Methods The human skin fibroblast HS27 cell line was used to examine cell viability and collagen synthesis after potassium chloride (KCl) t
This study proposes designing and applying a video stabilization algorithm using LIDAR-inertial odometry (LIO). The performance of the video stabilization function was validated in narrow, low-light environments using a small robot. A unique camera pose estimation method based on LIO has been introduced and integrated with the camera system in this work to not only overcome the limitations of image feature-based camera pose estimation but also address the issues arising from camera sensors. Thus
자율 주행 차량을 위한 정밀도로지도의 생성과 갱신의 요구가 증대함에 따라 구축과 운용 비용이 저렴한 다중 카메라 시스템이 측량 장비의 센서로 활용되고 있다. 다시점 영상의 스테레오 정합과 특징점 매칭 및 이를 통한 측위 등의 기술이 이러한 카메라 기반 3차원 지도 복원에 활용되고 있다. 본 논문에서는 차량의 상부에 장착된 다중 카메라 시스템을 기반으로, 점밀도와 정확도가 높은 노면의 3차원 점군 복원을 위한 hexgrid 기반 노면 모델 방법과 다시점 영상을 기반의 키프레임 자세와 3차원 점군을 이용하여 높이를 추정하고 텍스쳐를 정합하는 방법론을 제안한다. 초광각 어안렌즈가 장착된 다중 카메라와 GPS를 장착한 측량 시스템에 적용한 결과 최소 점 간격 0.025 m의 정밀한 도로 모델 생성이 가능함을 확인하였다.
In this paper, we present a framework for non-rigid alignment between a canonical 3D mesh of a non-rigid object and a single RGB-D image in a category-agnostic manner. Conventional optimization-based methods typically rely solely on geometric features, making them highly sensitive to ambiguities caused by occlusions and partial observations. To address these limitations, we propose the Neural Descriptor Field (NDF), which leverages semantic and geometric representations of pre-trained foundation
넓은 화각을 가진 어안렌즈는 한 번에 넓은 영역을 인식할 수 있어 자동차나 로봇 등 이동 플랫폼의 센서로 다양하게 사용되고 있다. 여러 대의 어안렌즈 카메라를 이용하여 360도 전방향의 거리와 플랫폼의 움직임을 추정할 수 있는 알고리즘도 개발되어 로봇 센서, 3차원 모델링 등에 활용되고 있다. 본 논문에서는 화각 220도 이상의 초광각 어안렌즈를 위한 새로운 투영 모델과, 이 렌즈를 장착한 멀티카메라 시스템의 카메라 내부 파라미터 캘리브레이션 및 각 카메라의 회전과 위치를 나타내는 외부 파라미터를 빠르고 간단하게 추정할 수 있는 카메라 시스템 캘리브레이션 방법론을 제안한다. 비교적 넓은 베이스라인의 초광각 다중 카메라 시스템의 캘리브레이션에 적용한 결과 서브픽셀 수준의 정확한 캘리브레이션이 가능함을 확인하였다.
Research Areas
Dive deeper into Jong Woo Im's research on Nubint
Open this lab's papers in the app to read with AI, summarize, and cite in your writing.