William June Suk Choi

About

I am a second-year Ph.D. student at Korea Advanced Institute of Science and Technology (KAIST), advised by Prof. Kimin Lee and Prof. Jinwoo Shin. I received a B.S. degree in computer science at KAIST. I am currently interested in developing capable and safe generative AI, focusing especially on effective and efficient generative modeling of images and videos with diffusion models.

Publications

Selected
All

[P2] Enhancing Motion Dynamics of Image-to-Video Models via Adaptive Low-Pass Guidance

June Suk Choi, Kyungmin Lee, Sihyun Yu, Yisol Choi, Jinwoo Shin, Kimin Lee

Preprint

arXiv Website Code

[C5] Enhancing LLM Agent Safety via Causal Influence Prompting

Dongyoon Hahm, Woogyeol Jin, June Suk Choi, Sungsoo Ahn, Kimin Lee

Findings of the Association for Computational Linguistics (ACL Findings), 2025

arXiv Code

[C4] Silent Branding Attack: Trigger-free Data Poisoning Attack on Text-to-Image Diffusion Models

Sangwon Jang, June Suk Choi, Jaehyeong Jo, Kimin Lee^{^}, Sung Ju Hwang^{^}

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025

arXiv Website Code

[C3] DiffusionGuard: A Robust Defense Against Malicious Diffusion-based Image Editing

June Suk Choi, Kyungmin Lee, Jongheon Jeong, Saining Xie, Jinwoo Shin, Kimin Lee

International Conference on Learning Representations (ICLR), 2025

arXiv Website Code

[C2] DiffExp: Efficient Exploration in Reward Fine-tuning for Text-to-Image Diffusion Models

Daewon Chae^*, June Suk Choi^*, Jinkyu Kim^{^}, Kimin Lee^{^}

AAAI Conference on Artificial Intelligence (AAAI), 2025(Oral Presentation, 547/12957=4.2%)

arXiv

[C1] Collaborative Score Distillation for Consistent Visual Editing

Subin Kim^*, Kyungmin Lee^*, June Suk Choi, Jongheon Jeong, Kihyuk Sohn, Jinwoo Shin

Conference on Neural Information Processing Systems (NeurIPS), 2023

arXiv Website Code

[P1] MobileSafetyBench: Evaluating Safety of Autonomous Agents in Mobile Device Control

Juyong Lee^*, Dongyoon Hahm^*, June Suk Choi^*, W. Bradley Knox, Kimin Lee

Preprint

arXiv Website Code

[P2] Enhancing Motion Dynamics of Image-to-Video Models via Adaptive Low-Pass Guidance

June Suk Choi, Kyungmin Lee, Sihyun Yu, Yisol Choi, Jinwoo Shin, Kimin Lee

Preprint

Arxiv Website Code

[C5] Enhancing LLM Agent Safety via Causal Influence Prompting

Dongyoon Hahm, Woogyeol Jin, June Suk Choi, Sungsoo Ahn, Kimin Lee

Findings of the Association for Computational Linguistics (ACL Findings), 2025

Arxiv Code

[C4] Silent Branding Attack: Trigger-free Data Poisoning Attack on Text-to-Image Diffusion Models

Sangwon Jang, June Suk Choi, Jaehyeong Jo, Kimin Lee^{^}, Sung Ju Hwang^{^}

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025

Arxiv Website Code

[C3] DiffusionGuard: A Robust Defense Against Malicious Diffusion-based Image Editing

June Suk Choi, Kyungmin Lee, Jongheon Jeong, Saining Xie, Jinwoo Shin, Kimin Lee

International Conference on Learning Representations (ICLR), 2025

Arxiv Website Code

[C2] DiffExp: Efficient Exploration in Reward Fine-tuning for Text-to-Image Diffusion Models

Daewon Chae^*, June Suk Choi^*, Jinkyu Kim^{^}, Kimin Lee^{^}

AAAI Conference on Artificial Intelligence (AAAI), 2025(Oral Presentation, 547/12957=4.2%)

Arxiv

[C1] Collaborative Score Distillation for Consistent Visual Editing

Subin Kim^*, Kyungmin Lee^*, June Suk Choi, Jongheon Jeong, Kihyuk Sohn, Jinwoo Shin

Conference on Neural Information Processing Systems (NeurIPS), 2023

Arxiv Website Code

[P1] MobileSafetyBench: Evaluating Safety of Autonomous Agents in Mobile Device Control

Juyong Lee^*, Dongyoon Hahm^*, June Suk Choi^*, W. Bradley Knox, Kimin Lee

Preprint

Arxiv Website Code

Education

Korea Advanced Institute of Science and Technology (KAIST)Sep. 2023 - Present

Ph.D. Student in Artificial Intelligence

Korea Advanced Institute of Science and Technology (KAIST)Mar. 2019 - Aug. 2023

B.S. in Computer Science (summa cum laude)

Experience

Chief Technology Officer and Co-founderJan. 2022 - Jan. 2023

PLUME.D

Proposed and prototyped AvaKit, the company's core product, securing ₩1B (≈$700K) in private and government funding while serving as the CTO.
1st place (of 876 teams), 11th Chung Ju-Young Startup Competition (largest private startup contest in Korea).
2nd place (of 5,420 teams), K-Startup Grand Challenge 2022 (largest government startup contest in Korea).

Academic Services

Journal Reviewer, TMLR

Honors & Awards

Dean's List, KAIST College of Engineering (2020F)Jan. 2021

Recipient, National Science and Engineering Undergraduate ScholarshipMay. 2020