Hyunsu Kim

Research engineer at NAVER LABS

gustnxodjs@gmail.com

About

I am a 3D computer vision research engineer on the Spatial AI team at NAVER LABS, where I turn real-world captures into detailed, large-scale 3D reconstructions of actual places. I work across the full pipeline — squeezing every bit of quality and scale out of 3D reconstruction (Gaussian Splatting) on raw, in-the-wild data, then refining the result with generative models. What drives me most is owning this path end to end, from raw captures to high-quality 3D.

Previously, I worked as a software engineer (ML R&D) on the XR team at Ohouse, specializing in 3D vision and generative models. I played a key role in developing AR solutions such as the 'Space Saving' and 'Furniture Removal' features, which enable users to visualize their space with accurately scaled 3D models and experiment with virtual furniture placements. I worked closely with Hyowon Ha and Jaeheung Surh to enhance user experiences in extended reality. Before that, I was a research scientist on the generation research team at NAVER AI Lab, collaborating with Junho Kim, Gayoung Lee, and Yunjey Choi, as well as Prof. Jun-Yan Zhu (CMU). I received my B.S. and M.S degrees in Computer Science and Engineering from Seoul National University in 2019 and 2021, respectively, where I was advised by Prof. Sungjoo Yoo during my Master’s program.

Projects in Ohouse: "Space Saving" & "Furniture Removal"

When purchasing furniture, do you wonder if it will match your space or if the size is right? Try out these new features!

✨ "Space Saving" Feature:

Capture your space, and it will be converted into a 3D model with accurate dimensions. You can virtually place new furniture to see how it fits.

✨ "Furniture Removal" Feature:

Remove existing furniture virtually without physically moving it and visualize your space with new furniture in place.

🔗 Explore AR-compatible furniture

🔗 Meet the Ohouse XR Team

📹 Capture:
🚀 Demo:

Publications

Most recent publications on Google Scholar.
‡ indicates equal contribution.

GLINT: Modeling Scene-Scale Transparency via Gaussian Radiance Transport

Youngju Na, Jaeseong Yun, Soohyun Ryu, Hyunsu Kim, Sung-Eui Yoon, Suyong Yeon

CVPR, 2026 (Oral, Award Candidate)

Project Paper Code

InSpace: Structure-Aware 3D Indoor Scene Generation from a Single 360° Image

Gwanhyeong Koo, Hyunsu Kim, Youngji Kim, Taejae Lee, Siwoo Lim, Sunjae Yoon, Suyong Yeon‡, Chang D. Yoo‡

Under review

LighthouseGS: Indoor Structure-aware 3D Gaussian Splatting for Panorama-Style Mobile Captures

Seungoh Han, Jaehoon Jang, Hyunsu Kim, Jaeheung Surh, Hyowon Ha‡, Kyungdon Joo‡

WACV, 2026

Project Paper

3D-aware Blending with Generative NeRFs

Hyunsu Kim, Gayoung Lee, Yunjey Choi, Jin-Hwa Kim, Jun-Yan Zhu

ICCV, 2023

Project Paper Code

BallGAN: 3D-aware Image Synthesis with a Spherical Background

Minjung Shin, Yunji Seo, Jeongmin Bae, Young Sun Choi, Hyunsu Kim, Hyeran Byun, Youngjung Uh

ICCV, 2023

Project Paper Code

Diffusion Models with Grouped Latents for Interpretable Latent Space

Sangyun Lee, Gayoung Lee, Hyunsu Kim, Junho Kim, Youngjung Uh

ICMLW, 2023

Paper

User-friendly Image Editing with Minimal Text Input: Leveraging Captioning and Injection Techniques

Sunwoo Kim, Wooseok Jang, Hyunsu Kim, Junho Kim, Yunjey Choi, Seungryong Kim, Gayeong Lee

arXiv, 2023

Paper

Diffusion Video Autoencoders: Toward Temporally Consistent Face Video Editing via Disentangled Video Encoding

Gyeongman Kim, Hajin Shim, Hyunsu Kim, Yunjey Choi, Junho Kim, Eunho Yang

CVPR, 2023

Project Paper Code

Context-Preserving Two-Stage Video Domain Translation for Portrait Stylization

Doyeon Kim, Eunji Ko, Hyunsu Kim, Yunji Kim, Junho Kim, Dongchan Min, Junmo Kim, Sung Ju Hwang

CVPRW, 2023

Paper

Learning Input-agnostic Manipulation Directions in StyleGAN with Text Guidance

Yoonjeon Kim, Hyunsu Kim, Junho Kim, Yunjey Choi, Eunho Yang

ICLR 2023

Paper Code

Generator Knows What Discriminator Should Learn in Unconditional GANs

Gayoung Lee, Hyunsu Kim, Junho Kim, Seonghyeon Kim, Jung-Woo Ha, Yunjey Choi

ECCV 2022

Paper Code

Generating videos with dynamics-aware implicit generative adversarial networks

Sihyun Yu‡, Jihoon Tack‡, Sangwoo Mo‡, Hyunsu Kim, Junho Kim, Jung-Woo Ha, Jinwoo Shin

ICLR 2022

Project Paper Code

StyleMapGAN: Exploiting Spatial Dimensions of Latent in GAN for Real-time Image Editing

Hyunsu Kim, Yunjey Choi, Junho Kim, Sungjoo Yoo, Youngjung Uh

CVPR 2021

Paper Video Code

Tag2pix: Line art colorization using text tag with secat and changing loss

Hyunsu Kim‡, Ho Young Jhoo‡, Eunhyeok Park, Sungjoo Yoo

ICCV 2019

Paper Code

Vitæ

Template originally from Martin Saveski