Keon Hee Park

I am a Ph.D. student, working on continual learning, domain adaptation, and category discovery for robust visual recognition under distribution shifts.

📧 Email: keonhee.park@vision.snu.ac.kr

Publications

[8] Detecting Unknown Objects via Energy-based Separation for Open World Object Detection

J.-W. Heo*, Keonhee Park*, G.-M. Park
CVPR 2026, Denver, USA

[7] SFUOD: Source-Free Unknown Object Detection

Keonhee Park*, S.-A. Choe, G.-M. Park
ICCV 2025, Hawaii, USA

[6] Universal Domain Adaptation for Semantic Segmentation

S.-A. Choe*, Keonhee Park, Jinwoo Choi, G.-M. Park
CVPR 2025, Nashville, USA

[5] CED: Comparing Embedding Differences for Detecting Out-of-Distribution and Hallucinated Text

Hakyung Lee*, Keonhee Park, Hoyoon Byun, Jeyoon Yeom, Jihee Kim, G.-M. Park, and Kyungwoo Song
EMNLP Findings 2024, Miami, USA

[4] Online Continuous Generalized Category Discovery

Keonhee Park*, Hakyung Lee, Kyungwoo Song, G.-M. Park
ECCV 2024, Milan, Italy

[3] Pre-trained Vision and Language Transformers Are Few-Shot Incremental Learners

Keonhee Park*, Kyungwoo Song, G.-M. Park
CVPR 2024, Seattle, USA

[2] Open Set Domain Adaptation for Semantic Segmentation

S.-A. Choe*, A.-H. Shin*, Keonhee Park, Jinwoo Choi, and G.-M. Park
CVPR 2024, Seattle, USA

[1] Online Class Incremental Learning on Stochastic Blurry Task Boundary via Mask and Visual Prompt Tuning

J.-Y. Moon*, Keonhee Park*, J.U. Kim, and G.-M. Park
ICCV 2023, Paris, France