Carview!

CARVIEW

MOTORHOMES

Select Language

HTTP/2 200 server: GitHub.com content-type: text/html; charset=utf-8 last-modified: Tue, 16 Dec 2025 14:01:47 GMT access-control-allow-origin: * strict-transport-security: max-age=31556952 etag: W/"6941664b-50f5" expires: Tue, 30 Dec 2025 18:14:09 GMT cache-control: max-age=600 content-encoding: gzip x-proxy-cache: MISS x-github-request-id: A5F0:234FE9:A73EF3:BBA15D:69541417 accept-ranges: bytes age: 0 date: Tue, 30 Dec 2025 18:04:09 GMT via: 1.1 varnish x-served-by: cache-bom-vanm7210053-BOM x-cache: MISS x-cache-hits: 0 x-timer: S1767117850.688452,VS0,VE204 vary: Accept-Encoding x-fastly-request-id: bb262dc8f011bcfc06f3067484b99286360af3ab content-length: 6196 Daehyeon Choi · Daehyeon Choi

Daehyeon Choi

Master Student in KAIST CS
KAIST Visual AI Group
daehyeonchoi[at]kaist.ac.kr

About Me

I am Daehyeon Choi, First-year Master student in KAIST, starting from Spring, 2025. I am a member of KAIST Visual AI Group, advised by Professor Minhyuk Sung.
Previously, I recevied a Bachelor of Science (B.S.) degree at POSTECH, 2025. I was an Intern in GenGenAI and Meissa Planet. And, I was an undergraduate intern in SNU, POSTECH, and KAIST.
Currently, I am interested in Large Vision-Language Model, especially Visual Grounding leveraging pre-trained large-scale model's prior.

Contact Info:
Email / Linkedin / CV / Google Scholar

Articles:
Reporter of The Postech Times

Research Interests

Vision-Language Model:
Embodied Question Answering, Vision-Language-Action, Spatial Reasoning
Generative Modeling:
Image Synthesis, Image Editing, Generative PDE

Recent News

(Dec. 12th, 2025) A paper about Ambulatory Vision System is released.
(Dec. 1st, 2025) I won the National Master's Student Excellence Scholarship.
(Sep. 18th, 2025) ORIGEN is accepted to NeurIPS 2025. See you in San Diego!
(Apr. 1st, 2025) ORIGEN is released on Arxiv.
(Feb. 7th, 2025) I earned B.S. degree from POSTECH.
(Dec. 2nd, 2024) I won Donggyo Talent Fellowship.
(Oct. 30th, 2024) I will start M.S. at KAIST with Prof. Minhyuk Sung.
(Sep. 12th, 2024) InstantDrag is accepted to SIGGRAPH Asia 2024.
(Jun. 24th, 2024) Intern at GenGenAI.

Publications

Toward Ambulatory Vision: Learning Visually-Grounded Active View Selection (Preprint)

Juil Koo*, Daehyeon Choi*, Sangwoo Youn*, Phillip Y. Lee, Minhyuk Sung (* Equal Contribution)

Unlike Snapshot Vision, which answers queries solely from given observations, we introduce Ambulatory Vision, searching for visual clues in the current scene and moving to acquire more informative observations.

Project Page / Code / Arxiv

ORIGEN: Zero-shot 3D Orientation Grounding in Text-to-Image Generation (NeurIPS 2025)

Yunhong Min*, Daehyeon Choi*, Kyeongmin Yeo, Jihyun Lee, Minhyuk Sung (* Equal Contribution)

We introduce ORIGEN, the first zero-shot method for 3D orientation grounding in text-to-image generation.

Project Page / Code / Arxiv

InstantDrag: Improving Interactivity in Drag-based Image Editing (SIGGRAPH Asia 2024)

Joonghyuk Shin, Daehyeon Choi, Jaesik Park

InstantDrag performs drag-edits in about a second requiring minimal inputs and memory.

Project Page / Code / Arxiv

Education

Korea Advanced Institute of Science and Technology (KAIST), (Spring, 2025 - Current)
- Master Student, Dept. of Computer Science
- Graduate Student in KAIST Visual AI Group, KAIST

Seoul National University, (Spring, 2024)
- Exchange Student, Dept. of Electronic & Computer Engineering
- Research Intern in Visual & Geometric Intelligence Lab, SNU

Pohang Institute of Science & Technology (POSTECH), (2021 - 2024)
- Bachelor of Science (B.S.), Dept. of Computer Science & Engineering (Cumulative GPA: 3.85/4.30)

Jeonbuk Scicence High School, (2019 - 2020)
- Early Graduation in 4 semesters

Work Experiences

Algorithm Engineer Intern, GenGenAI. (Jun. 2024 ~ Aug. 2024)
- Verification and analysis of synthetic-data based training (Multi-class classification)
- Based on Stable Diffusion, Synthetic data generation in specific case

Research Intern, Meissa Planet. (Jun. 2023 ~ Aug. 2023)
- In Research Team, Satelite Analysis with Deep learning, especially Semantic Change Detection

Research Experiences

Undergraduate Research Intern, Visual AI Group, KAIST. (Jun. 2024 ~ Aug. 2024)
- Diffusion Models, 3D-aware Editing
- Under the supervision by Prof. Minhyuk Sung

Undergraduate Research Intern, VGI Lab, SNU. (Oct. 2023 ~ Jun. 2024)
- Diffusion Models, Drag Editing
- Participated to InstantDrag (SIGGRAPH Asia 2024).
- Under the supervision by Prof. Jaesik Park, co-worked with Joonghyuk Shin.

Undergraduate Research Intern, CV Lab, POSTECH. (Sep. 2022 ~ Feb. 2023)
- Generative Adversarial Networks
- Under the supervision by Prof. Jaesik Park

Award & Scholarships

National Master's Student Excellence Scholarship, Korea Government. (Dec. 2025)
- 2,500,000 KRW per semester
- Outstanding master students majoring Science and Engineering in Korea.

Donggyo Talent Fellowship, Silver Prize, Surim Foundation. (Dec. 2024)
- 1,000,000 KRW
- Outstanding 25 undergrads who have greatly contributed to the promotion and development of the university's honor.

In Jae Rim Scholarships, Korea Foundation For Advanced Studies (KFAS). (Jan. 2024 ~ Current)
- 4,000,000 KRW per semester
- Completed programs: Imagine to Change, SK FLP Mentoring, Conflict Resolution, SOUL Project.

Tae–Jun Park Alumni Scholarships, POSTECH Alumni Foundation. (Oct. 2023)
- 3,000,000 KRW

Global Leadership Program Fellowship, CSE, POSTECH. (2022F, 2023S, 2023F)
- (2022F, 2023S) 3,000,000 KRW
- (2023F) 8,000,000 KRW at maximum

Grand Prize(1st), Undergraduate Group Research Program, POSTECH. (Feb. 2023)
- 4,000,000 KRW

STX Scholarships, STX Foundation, (Jan. 2023 ~ Current)
- 500,000 KRW per month & All tuition Fees

Jigok Scholarships, POSTECH, (Feb. 2021 ~ Dec. 2022)
- All tuition Fees

Academic Services

Journal Reviewer
- TMLR

Original Source | Taken Source

Daehyeon Choi Master. Student @ KAIST, C.S

Daehyeon Choi

About Me

Research Interests

Recent News

Publications

Education

Work Experiences

Research Experiences

Award & Scholarships

Academic Services