Hyewon Park | Ewha W. University

Biography

Passionate researcher in computer vision, with a focus on unsupervised learning and out-of-domain generalization. Experienced in developing parameter-efficient fine-tuning (PEFT) methods for continual learning and test-time adaptation (TTA).

Received an M.S. in Computer Science from Ewha W. University, advised by Prof. Dongbo Min in Computer Vision Lab (CVLab). Conducted research on robust adaptation techniques to improve model performance under domain shifts. Previously worked for 2 years as a Data Scientist at an AI consulting firm, gaining diverse industry experience across various projects.

Committed to advancing robust adaptation methodologies to enhance model reliability under domain shifts.

Research Interests

Self-supervised learning
Out-of-Domain Generalization
Parameter-efficient Fine-Tuning
Scene Understanding

News

[Sep. 2025] One paper got accepted in NeurIPS 2025!
[Jun. 2025] One paper got accepted in ICCV 2025🏝️!

Publications

NeurIPS

RobIA: Robust Instance-aware Continual Test-time Adaptation for Deep Stereo

Jueun Ko*, Hyewon Park*, Hyesong Choi, Dongbo Min (*: Equal Contribution)

Neural Information Processing Systems (NeurIPS), 2025.

PDF Code

ICCV

⛵Hybrid-TTA: Continual Test-time Adaptation via Dynamic Domain Shift Detection

Hyewon Park, Hyejin Park, Jueun Ko, Dongbo Min

International Conference of Computer Vision (ICCV), 2025.

PDF Code