Jaehoon Jang

I'm a M.S student in 3D Vision & Robotics lab at UNIST, supervised by Prof. Kyungdon Joo. I am broadly interested in 3D computer vision and deep learning. Recently, my research focuses on creating 3D contents with neural rendering, implicit neural representations, and generative models.

I am now a visiting scholar at Visual Computing Group, Harvard, supervised by Prof. Hanspeter Pfister, and working closely with Dr. Wanhua Li.

Email / LinkedIn / Github

Publications

	Practical 3D Gaussian Splatting for Indoor Scenes Under review, Co-first author
	Physical Interaction-Guided 3D Human Generation Under review
	AiSDF: Structure-aware Neural Signed Distance Fields in Indoor Scenes Jaehoon Jang, Inha Lee, Minje Kim, Kyungdon Joo RA-L, 2024 [ Project page \| arXiv ] Given a stream of posed depth images, we propose a structure-aware online SDF reconstruction framework for indoor scenes under the Atlanta world assumption. (* Equal contribution. Alphabetic order)
	ContactGen: Contact-Guided Interactive 3D Human Generation for Partners Dongjun Gu, Jaehyeok Shim, Jaehoon Jang, Changwoo Kang, Kyungdon Joo AAAI, 2024 [ Project page \| arXiv ] Generating a interactive 3D human avatar considering physical contact.

Awards

Winner of AI Technology Open Workshop, UNIST, 2024 (Gifted by HD Hyundai)

Education

M.S in Artificial Intelligence Graduate School, UNIST (2022.03 ~ )

Bachelor in Automobile engineering, Kookmin Univ (2016.03. ~ 2022.02)

Projects

Novel view synthesis, funded by Bucketplace (2023.09 ~ )

Commonsense based 3D human pose estimation, funded by IITP (2022.03 ~ )

Template from this website.