I'm a postdoc at Cornell Tech, working with Prof. Hadar Averbuch-Elor. I am interested in building machines that can understand the rich spatio-temporal contexts in our world, and ultimately contribute to advances in science and humanities.
Sep 2026
Started my new position as postdoc at Cornell Tech 🗽.
Jun 2026
One paper is accepted to IROS 2026.
Mar 2026
One paper is accepted to CVPR 2026 Findings.
Feb 2026
One paper is accepted to IEEE Transactions on Robotics (T-RO).
Point2Act is accepted to ICRA 2026.
I will be giving a talk at the OmniCV workshop in CVPR 2026.
Oct 2025
Event-based localization received the best poster award at the ICCV 2025 workshop on "Responsible Imaging"
A new preprint on modeling long-term 3D scene changes is out: Long-Term Gaussian Scene Chronology.
Sep 2025
I've been selected for participation at the ICCV 2025 doctoral consortium.
Event-based localization is accepted to ICCV 2025 workshop on "Responsible Imaging"
Point2Act is accepted to CoRL 2025 workshop on "Generalizable Priors for Robot Manipulation"
3D scene analogies is accepted to CoRL 2025 workshop on "Learning Effective Abstractions for Planning (LEAP)" and ICCV 2025 workshop on "Open-World 3D Scene Understanding with Foundation Models".
Analogical Trajectory Transfer
Junho Kim, Eunsun Lee, Gwangtak Bae, Seunggu Kang, Young Min Kim
arXiv preprint, 2026
[arXiv]
Using 3D foundation models, we can transfer trajectories from one scene to another while preserving the contextual information.
AnyCamVLA: Zero-Shot Camera Adaptation for Viewpoint Robust Vision-Language-Action Models
Hyeongjun Heo, Seungyeon Woo, Sang Min Kim, Junho Kim, Junho Lee, Yonghyeon Lee, Young Min Kim
IROS 2026
[arXiv]
To make pre-trained VLAs generalizable to viewpoint changes, we propose using feed-forward novel view synthesis which doesn't require additional VLA fine-tuning or demo collection.
LTGS: Long-Term Gaussian Scene Chronology From Sparse View Updates
Minkwan Kim, Seungmin Lee, Junho Kim, Young Min Kim
CVPR 2026 (Findings)
[arXiv] [Code] [Project Page]
We design a light-weight pipeline to estimate the long-term changes within scenes from spatio-temporally sparse images.
RoEL: Robust Event-based 3D Line Reconstruction
Gwangtak Bae, Jaeho Shin, Seunggu Kang, Junho Kim, Ayoung Kim, Young Min Kim
IEEE Transactions on Robotics (T-RO) 2026
[arXiv] [Code] [Project Page]
From posed event streams, we detect and aggregate 2D lines to produce a 3D line map, which can be further used for cross-modal applications.
Point2Act: Efficient 3D Distillation of Multimodal LLMs for Zero-Shot Context-Aware Grasping
Sang Min Kim, Hyeongjun Heo, Junho Kim, Yonghyeon Lee, Young Min Kim
ICRA 2026, CoRL 2025 Workshop on Generalizable Priors for Robot Manipulation
[arXiv] [Code] [Project Page]
By combining multimodal foundation models with neural radiance fields, we demonstrate efficient zero-shot grasping from text prompts.
Privacy-Preserving Visual Localization with Event Cameras
Junho Kim, Young Min Kim, Ramzi Zahreddine, Weston Anthony Welge, Gurunandan Krishnan, Sizhuo Ma, Jian WangÂ
IEEE Transactions on Image Processing (TIP) 2025, ICCV 2025 Workshop on Responsible Imaging (Best Poster Award, Oral Presentation)
[arXiv] [Code] [Paper] [Video] [Project Page]
We propose an event-based visual localization method that effectively leverages the strengths of event cameras while offering privacy protection for alleviating user concerns in client-server localization scenarios.Â
Finding 3D Scene Analogies with Multimodal Foundation Models
Junho Kim, Young Min Kim
RSS 2025Â Workshop on Large Foundation Models for Interactive Robot Learning (Spotlight Presentation)
[arXiv]
We leverage multimodal foundation models to find 3D scene analogies in complex indoor scenes.
Learning 3D Scene Analogies with Neural Contextual Scene Maps
Junho Kim, Gwangtak Bae, Eunsun Lee, Young Min Kim
ICCV 2025, CoRL 2025 Workshop on Learning Effective Abstractions for Planning, ICCV 2025 Workshop on Open-World 3D Scene Understanding with Foundation Models
[arXiv] [Code] [Paper] [Video] [Project Page]
We propose a new task of finding 3D scene analogies, which are dense maps connecting regions sharing similar scene contexts.
Fully Geometric Panoramic Localization
Junho Kim, Jiwon Jeong, Young Min Kim
CVPR 2024Â (Nectar Track Spotlight Presentation at 3DV 2025)
[arXiv] [Code] [Paper] [Video] [Project Page]
By leveraging lines and their intersections, we can perform panoramic localization without using any visual features, in a fully geometric manner.
LDL: Line Distance Functions for Panoramic Localization
Junho Kim, Changwoon Choi, Hojun Jang, Young Min Kim
ICCV 2023
[arXiv] [Code] [Paper] [Video] [Project Page]
We introduce a simple pose search method that operates on 2D, 3D line segments based on distance functions, which attains competitive performance with a very short runtime.
Calibrating Panoramic Depth Estimation for Practical Localization and Mapping
Junho Kim, Eunsun Lee, Young Min Kim
ICCV 2023
[arXiv] [Code] [Paper] [Video] [Project Page]
With a simple test-time adaptation scheme, we can 'calibrate' panoramic depth estimation algorithms to make more robust predictions and further improve on downstream tasks in localization and mapping.
Ev-NeRF:Â Event Based Neural Radiance Field
Inwoo Hwang, Junho Kim, and Young Min Kim
WACV 2023
[arXiv] [Paper] [Video] [Project Page]
Neural radiance fields (NeRF) for event cameras can be learned without using any intensity image supervision, and enables robust high-dynamic range imaging amidst large amounts of sensor noise.
CPO: Change Robust Panorama to Point Cloud Localization
Junho Kim, Hojun Jang, Changwoon Choi, and Young Min Kim
ECCV 2022
[arXiv] [Code] [Paper] [Video] [Project Page]
Constructing score maps in 2D, 3D that reflect regional color consistencies enable robust localization amidst scene changes.
MoDA:Â Map style transfer for self-supervised Domain Adaptation of embodied agents
Eunsun Lee, Junho Kim, Sangwon Park, and Young Min Kim
ECCV 2022
[arXiv] [Paper] [Video] [Project Page]
By applying style transfer on the grid map domain, effective domain adaptation is possible for environments containing visual or dynamic noise.
Ev-TTA: Test-Time Adaptation for Event-Based Object Recognition
Junho Kim, Inwoo Hwang, and Young Min Kim
CVPR 2022
[arXiv] [Code] [Paper] [Video] [Project Page]
A simple test-time adaptation objective can help event-based classifiers to make robust predictions in extreme conditions such as large motion or low lighting.Â
Self-Supervised Domain Adaptation for Visual Navigation with Global Map Consistency
Eunsun Lee, Junho Kim, and Young Min Kim
WACV 2022Â
[arXiv] [Paper] [Video] [Project Page]
Imposing global map consistency by performing round-trips can enhance the navigation performance of embodied agents amidst various dynamics corruptions.
SGoLAM: Simultaneous Goal Localization and Mapping for Multi-Object Goal Navigation
Junho Kim, Eunsun Lee, Mingi Lee, Dongsu Zhang, and Young Min Kim
arXiv preprint, 2021. (2nd place in the MultiON challenge @ Embodied AI workshop in CVPR 2021)Â
[arXiv] [Code] [Video]
By performing goal localization and mapping simultaneously with a simple color matching scheme, embodied agents can effectively perform multi-object goal navigation.
PICCOLO: Point Cloud-Centric Omnidirectional Localization
Junho Kim, Changwoon Choi, Hojun Jang, and Young Min Kim
ICCV 2021Â
[arXiv] [Code] [Paper] [Video] [Project Page]
By minimizing a novel loss function that penalizes color discrepancies in 2D and 3D, effective localization can be performed using panorama images without learning.
N-ImageNet: Towards Robust, Fine-Grained Object Recognition with Event Cameras
Junho Kim, Jaehyeok Bae, Gangin Park, Dongsu Zhang, and Young Min Kim
ICCV 2021Â
[arXiv] [Code] [Paper] [Video] [Project Page]
We propose a large scale dataset for event-based object recognition that enables systematic quantification of classification robustness amidst extreme lighting or motion.
Meta Reality Labs, Seattle, Washington, May 2023 - Aug. 2023
Research intern, XR Spatial AI Group
Mentors: Abduallah Mohamed, True Price, and Fede Camposeco
Snap Research, New York City, New York, May 2022 - Sep. 2022
Research intern, Computational Imaging Group
Mentors: Sizhuo Ma, Jian Wang, and Yicheng Wu
Seoul National University, Seoul, Korea, Mar. 2020 - Present
M.S./Ph.D. in Electrical and Computer Engineering
Seoul National University, Seoul, Korea, Mar. 2016 - Feb. 2020
B.S. in Electrical and Computer Engineering, Summa Cum Laude
KFAS (Korea Foundation for Advanced Studies) Graduate Study Scholarship, Sep. 2020 - Present
National Science and Engineering Scholarship, Mar. 2018 - Mar. 2020
Merit - Based Scholarship, Sep. 2016 - Mar. 2018
Reviewer: 3DV, ECCV, CVPR, WACV
Holistic Perception: From Seeing All Around to Analogical Reasoning, OmniCV workshop at CVPR 2026, Jun. 4 2026. [Slides]
On Two Pillars of Spatial Perception: Metric and Analogical Reasoning, Cornell Tech, Nov. 7 2025. [Slides]
Towards Actionable Spatial Perception, KCCV Doctoral Consortium, Aug. 4 2025. [Slides]
Panoramic Localization Using Line Geometry, AIST-JRL, Jun. 5 2025. [Slides]
Fully Geometric Panoramic Localization, Nectar Track at 3DV 2025, Mar. 28 2025. [Slides]
On Two Pillars of Spatial Perception: Metric and Analogical Reasoning, Hanyang University, Mar. 24 2025. [Slides]
LDL: Line Distance Functions for Panoramic Localization, AIIS Retreat in Seoul National University, Nov. 17 2023. [Video] [Slides]
PICCOLO: Point-Cloud Centric Omnidirectional Localization, Map-based Localization for Autonomous Driving (MLAD) workshop at ECCV 2022, Oct. 23 2022. [Video] [Slides]
Robust Visual Recognition in Extreme Conditions Using Event Cameras, Imaging Science Seminar at Institute of New Media and Communications (INMC) in Seoul National University, Mar. 31 2022. [Slides]