Hyunyoung Jung

I am a Ph.D. student in the Department of Computer Science and Engineering (CSE) at Seoul National University, where I am fortunate to be under the supervision of Prof. Sungjoo Yoo. In the summers of 2022 and 2023, I completed internships at Meta Reality Labs. Prior to my doctoral studies, I obtained a B.S. in CSE from Seoul National University.

My research interests broadly lie in computational photography, 2D/3D vision, and graphics. Within this area, I have experience with scene perception (such as depth, optical flow, and segmentation), visual localization, generative models (such as GANs and Diffusion models), and neural rendering.

CV / Email / Google Scholar / Linkedin

Research

	Geometry Transfer for Stylizing Radiance Fields Hyunyoung Jung, Seonghyeon Nam, Nikolaos Sarafianos, Sungjoo Yoo, Alexander Sorkine-Hornung, Rakesh Ranjan CVPR, 2024 paper / project page / arXiv The geometry and appearance of a 3D scene are coherently stylized, guided by an RGB image and a depth map.
	AnyFlow: Arbitrary Scale Optical Flow with Implicit Neural Representation Hyunyoung Jung, Zhuo Hui, Lei Luo, Haitao Yang , Feng Liu , Sungjoo Yoo, Rakesh Ranjan, Denis Demandolx CVPR, 2023 (Highlight) paper/ arXiv The optical flow network is designed to produce outputs at any desired resolution while maintaining robust performance, when processing low-resolution images.
	Fine-grained Semantics-aware Representation Enhancement for Self-supervised Monocular Depth Estimation Hyunyoung Jung, Eunhyeok Park, Sungjoo Yoo ICCV, 2021 (Oral) paper / arXiv / code The depth estimation network utilizes semantic information to enhance boundary accuracy, incorporating metric-learning and cross-attention.

Projects

	Reflectance-aware Generative Radiance Fields for 3D-aware Image Synthesis 2021 The generative NeRF-based network is trained to achieve relightability using only a collection of single-view images, without requiring any supplementary information.
	Development of model compression framework for scalable on-device AI computing on Edge applications A Korean government funded project ($10M, 2021-2024) to develop automatic DL model optimization methods for on-device AI on commercial neural network accelerators.

Experience

	Research Intern at Meta Reality Labs Jun 2023 - Dec 2023, Sunnyvale, CA, US Jun 2022 - Dec 2022, Seattle, WA, US
	Software Engineer Intern at Bobidi Jan 2022 - Feb 2022, Seoul, South Korea
	Software Engineer Intern at Line Plus Corp. Jan 2018 - Feb 2018, Seongnam, South Korea

Honors & awards

	Qualcomm Innovation Fellowship Korea 2021 Nov 2021
	NAVER LABS Mapping & Localization Challenge Jul 2020 Built full pipeline of structure-based hierarchical visual localization framework on NAVER LABS datasets. Earned 2nd Place in both Indoor / Outdoor sections (total prize 6M KRW).

Education

Seoul National University

Sep 2019 - Present
The Integrated MA/Ph.D. Course in Computer Science and Engineering

Mar 2013 - Aug 2019 (two gap years for military service)
Bachelor of Science in Computer Science and Engineering

Forked from Jon Barron's website .