Jiwon Kang
Integrated M.S./Ph.D. Student — KAIST AI (CVLAB)
loggerjk.github.io Google Scholar github.com/loggerJK jiwon7258@kaist.ac.kr LinkedIn

Education

KAIST, South Korea
M.S. in Artificial Intelligence, Kim Jaechul Graduate School of AI
Advisor: Prof. Seungryong Kim (CVLAB)
Mar. 2025 – Feb. 2027 (Expected)
Korea University, Seoul
B.S. in Computer Science and Engineering
Double Major in Statistics
GPA: 4.24 / 4.5
  • Leave of Absence for R.O.K. Military Service (Dec. 2020 – Jun. 2022)
  • Undergraduate Research Intern under Prof. Seungryong Kim (Sep. 2022 – Feb. 2025)
Mar. 2019 – Feb. 2025

Experience

Kakao, Seoul
Research Intern
Worked on Exploring Whisper as a Continuous Tokenizer for autoregressive diffusion-based Text-to-Speech
Mar. 2026 – May 2026

Publications

* denotes equal contribution.  ·  † denotes corresponding author.

International Conference

Transferability Between Understanding and Generation in Unified Multimodal Models
Jiwon Kang*, Heeji Yoon*, Jaewoo Jung, Jaewon Min, Minkyeong Jeon, Biyeon Hwang, Sangwon Jung, Seungryong Kim†
  • Investigated how understanding and generation capabilities (e.g., counting) transfer to one another within unified multimodal models.
  • Showed that transferability depends on the model architecture, and proposed a practical post-training strategy for unified multimodal models that leverages transferability.
ECCV, 2026
APPLE: Attribute-Preserving Pseudo-Labeling for Diffusion-Based Face Swapping
Jiwon Kang*, Yeji Choi*, JoungBin Lee, Wooseok Jang, Jinhyeok Choi, Taekeun Kang, Yongjae Park, Myungin Kim, Seungryong Kim†
  • Prior diffusion-based face swapping methods focused on changing facial identity, failing to preserve facial attributes.
  • Proposed a novel pseudo-labeling framework that preserves facial attributes while swapping identities, achieving Pareto-optimal results.
CVPR, 2026
A Noise is Worth Diffusion Guidance
Donghoon Ahn*, Jiwon Kang*, Sanghyun Lee, Jaewon Min, Wooseok Jang, Minjae Kim, Hyungwon Cho, Sayak Paul, SeonHwa Kim, Eunju Cha†, Kyong Hwan Jin†, Seungryong Kim†
  • Showed that classifier-free guidance can be replaced by refining the initial noise.
  • Enabled high-quality generation without a guidance pass, highlighting the importance of the initial noise in diffusion models.
ICLR, 2026
Where and How to Perturb: On the Design of Perturbation Guidance in Diffusion and Flow Models
Donghoon Ahn*, Jiwon Kang*, Sanghyun Lee, Minjae Kim, Wooseok Jang, Jaewon Min, Sangwu Lee, Sayak Paul, Seungryong Kim†
  • Systematically studied perturbation guidance in diffusion and flow models.
  • Showed that perturbing different layers/heads in a diffusion model exhibits different effects on the generated samples, which can be systematically combined to improve user-specified attributes via the proposed framework.
NeurIPS, 2025

Workshop

Self-Evolving Neural Radiance Fields
Jaewoo Jung*, Jisang Han*, Jiwon Kang*, Seongchan Kim, Min-Seop Kwak, Seungryong Kim†
  • Proposed a self-training framework that lets a NeRF improve itself from few input views via pseudo-label distillation.
ICCV Workshop (Wild3D), 2025

Preprints & Under Review

RAIN-GS: Relaxing Accurate Initialization Constraint for 3D Gaussian Splatting
Jaewoo Jung*, Jisang Han*, Honggyu An*, Jiwon Kang*, Seonghoon Park*, Seungryong Kim†
  • Relaxed the accurate point-cloud initialization requirement of 3D Gaussian Splatting, enabling training from sparse or random initialization.
Under Review

Extracurricular Activities

Military Service (KATUSA, Korean Augmentation to the United States Army)
194th DSSB, 2nd Infantry Division, Eighth U.S. Army, Camp Humphreys
Dec. 2020 – Jun. 2022
AIKU (Artificial Intelligence Society of Korea University)
Co-founder and Vice President — Academic research society for undergraduate students
Sep. 2022 – Aug. 2023