Donghyeon Kwon

My research focuses on learning with limited supervision and transferring knowledge across modalities, with applications in computer vision and 3D perception. I work on semi-supervised and self-supervised learning, as well as cross-modal knowledge distillation.

I received my Ph.D. in Computer Science and Engineering from POSTECH in 2026, advised by Prof. Suha Kwak in the Computer Vision Lab.

Military service obligation: I am completing my service as a researcher at GenGenAI, working on generative AI systems.

Portrait of Donghyeon Kwon
Taken in Prague, Czech Republic

Recent news

All news
  1. I received the Best Patent Award from the Samsung-POSTECH Research Center in 2026, together with Youngseok Yoon, Hyeongseok Son and Suha Kwak.

  2. Our papers on semi-supervised learning for LIDAR semantic segmentation and composed visual data retrieval are accepted to ECCV 2026.

  3. I received my Ph.D. in Computer Science and Engineering from POSTECH, where I conducted research in the Computer Vision Lab.

Publications

Most recent first
  1. Method overview: RePL: Pseudo-label Refinement for Semi-supervised LiDAR Semantic Segmentation

    RePL: Pseudo-label Refinement for Semi-supervised LiDAR Semantic Segmentation

    Donghyeon Kwon*, Taegyu Park*, Suha Kwak

    ECCV 2026 · * Equal contribution

    BibTeX
    @inproceedings{Kwon_2026_ECCV,
      title = {RePL: Pseudo-label Refinement for Semi-supervised LiDAR Semantic Segmentation},
      author = {Kwon*, Donghyeon and Park*, Taegyu and Kwak, Suha},
      booktitle = {Proceedings of the European Conference on Computer Vision (ECCV)},
      year = {2026},
      month = sep,
      pdf = {https://media.eventhosts.cc/Conferences/ECCV2026/pdfs/1997.pdf},
      note = {* Equal contribution},
      preview = {kwon_eccv_26.png},
      bibtex_show = {true}
    }
    
  2. Method overview: Learning Sample-wise Rank-Aware Interpolation Weights for Composed Visual Data Retrieval

    Learning Sample-wise Rank-Aware Interpolation Weights for Composed Visual Data Retrieval

    Boseung Jeong, Taegyu Park, Donghyeon Kwon, Hyunsouk Cho, Suha Kwak

    ECCV 2026

    BibTeX
    @inproceedings{Jung_2026_ECCV,
      title = {Learning Sample-wise Rank-Aware Interpolation Weights for Composed Visual Data Retrieval},
      author = {Jeong, Boseung and Park, Taegyu and Kwon, Donghyeon and Cho, Hyunsouk and Kwak, Suha},
      booktitle = {Proceedings of the European Conference on Computer Vision (ECCV)},
      year = {2026},
      month = sep,
      pdf = {https://media.eventhosts.cc/Conferences/ECCV2026/pdfs/5289.pdf},
      preview = {jung_eccv_26.png},
      bibtex_show = {true}
    }
    
  3. Method overview: MemDistill: Distilling LiDAR Knowledge into Memory for Camera-Only 3D Object Detection

    MemDistill: Distilling LiDAR Knowledge into Memory for Camera-Only 3D Object Detection

    Donghyeon Kwon, Youngseok Yoon, Hyeongseok Son, Suha Kwak

    ICCV 2025

    BibTeX
    @inproceedings{Kwon_2025_ICCV,
      title = {MemDistill: Distilling LiDAR Knowledge into Memory for Camera-Only 3D Object Detection},
      author = {Kwon, Donghyeon and Yoon, Youngseok and Son, Hyeongseok and Kwak, Suha},
      booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
      year = {2025},
      month = oct,
      pdf = {https://openaccess.thecvf.com/content/ICCV2025/papers/Kwon_MemDistill_Distilling_LiDAR_Knowledge_into_Memory_for_Camera-Only_3D_Object_ICCV_2025_paper.pdf},
      preview = {kwon_under25.png},
      bibtex_show = {true},
      pages = {6828-6838}
    }
    
  4. Method overview: Decoupled Finetuning for Domain Generalizable Semantic Segmentation

    Decoupled Finetuning for Domain Generalizable Semantic Segmentation

    Jaehyun Pahk, Donghyeon Kwon, Seong Joon Oh, Suha Kwak

    ICLR 2025

    BibTeX
    @inproceedings{pahk2025decoupled,
      title = {Decoupled Finetuning for Domain Generalizable Semantic Segmentation},
      author = {Pahk, Jaehyun and Kwon, Donghyeon and Oh, Seong Joon and Kwak, Suha},
      booktitle = {The Thirteenth International Conference on Learning Representations (ICLR)},
      year = {2025},
      month = apr,
      pdf = {https://openreview.net/forum?id=qZEdmyqCHF},
      preview = {pahk_iclr25.png},
      bibtex_show = {true}
    }
    
  5. Method overview: Boosting Semi-Supervised Video Action Detection with Temporal Context

    Boosting Semi-Supervised Video Action Detection with Temporal Context

    Donghyeon Kwon, Inho Kim, Suha Kwak

    WACV 2025 · Oral Presentation

    BibTeX
    @inproceedings{Kwon_2025_WACV,
      author = {Kwon, Donghyeon and Kim, Inho and Kwak, Suha},
      title = {Boosting Semi-Supervised Video Action Detection with Temporal Context},
      booktitle = {Proceedings of the Winter Conference on Applications of Computer Vision (WACV)},
      month = feb,
      year = {2025},
      note = {Oral Presentation},
      bibtex_show = {true},
      preview = {kwon_wacv25.png},
      pdf = {https://openaccess.thecvf.com/content/WACV2025/papers/Kwon_Boosting_Semi-Supervised_Video_Action_Detection_with_Temporal_Context_WACV_2025_paper.pdf}
    }
    
  6. Method overview: Self-Supervised Learning of Semantic Correspondence Using Web Videos

    Self-Supervised Learning of Semantic Correspondence Using Web Videos

    Donghyeon Kwon, Minsu Cho, Suha Kwak

    WACV 2024

    BibTeX
    @inproceedings{Kwon_2024_WACV,
      author = {Kwon, Donghyeon and Cho, Minsu and Kwak, Suha},
      title = {Self-Supervised Learning of Semantic Correspondence Using Web Videos},
      booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)},
      month = jan,
      year = {2024},
      pages = {2142-2152},
      bibtex_show = {true},
      html = {https://cvlab.postech.ac.kr/research/SSSCWEB/},
      pdf = {https://openaccess.thecvf.com/content/WACV2024/papers/Kwon_Self-Supervised_Learning_of_Semantic_Correspondence_Using_Web_Videos_WACV_2024_paper.pdf},
      preview = {kwon_wacv24.png}
    }
    
  7. Method overview: Leveraging Proxy of Training Data for Test-Time Adaptation

    Leveraging Proxy of Training Data for Test-Time Adaptation

    Juwon Kang, Nayeong Kim, Donghyeon Kwon, Jungseul Ok, Suha Kwak

    ICML 2023

    BibTeX
    @inproceedings{Kang_2023_ICML,
      author = {Kang, Juwon and Kim, Nayeong and Kwon, Donghyeon and Ok, Jungseul and Kwak, Suha},
      title = {Leveraging Proxy of Training Data for Test-Time Adaptation},
      booktitle = {International Conference on Machine Learning (ICML)},
      month = jul,
      year = {2023},
      bibtex_show = {true},
      html = {https://cvlab.postech.ac.kr/research/TTAwithProxy/},
      pdf = {https://openreview.net/pdf?id=VVGNInOAm9},
      preview = {kang_icml23.png}
    }
    
  8. Method overview: Semi-Supervised Semantic Segmentation With Error Localization Network

    Semi-Supervised Semantic Segmentation With Error Localization Network

    Donghyeon Kwon, Suha Kwak

    CVPR 2022

    BibTeX
    @inproceedings{Kwon_2022_CVPR,
      author = {Kwon, Donghyeon and Kwak, Suha},
      title = {Semi-Supervised Semantic Segmentation With Error Localization Network},
      booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
      month = jun,
      year = {2022},
      bibtex_show = {true},
      html = {https://cvlab.postech.ac.kr/research/ELN/},
      pdf = {https://openaccess.thecvf.com/content/CVPR2022/papers/Kwon_Semi-Supervised_Semantic_Segmentation_With_Error_Localization_Network_CVPR_2022_paper.pdf},
      preview = {kwon_cvpr22.png}
    }
    

Experience

  1. Mar 2026 – Feb 2027

    Researcher

    GenGenAI

    Military service obligation through February 28, 2027. Research and development on generative AI systems.

  2. Feb 2021 – Feb 2026

    Graduate Researcher

    Computer Vision Lab, POSTECH

    Advisor: Prof. Suha Kwak.

  3. Jul 2023 – Aug 2023

    Research Intern

    Samsung Advanced Institute of Technology

    Data-efficient 3D object detection for autonomous driving. Mentor: Dr. Hyeongseok Son.

  4. Jun 2019 – Aug 2019

    Research Assistant

    Korea Institute of Science and Technology

    Typo correction for virtual keyboards. Advisor: Dr. Suhyun Kim.

Education

  1. 2021 – 2026

    Ph.D. in Computer Science and Engineering

    POSTECH

    Advisor: Prof. Suha Kwak, Computer Vision Lab.

  2. 2017 – 2021

    B.S. in Computer Science and Engineering

    POSTECH

    Advisor: Prof. Hwanjo Yu.

Academic service

Conference reviewing

CVPR
2023–2026
ICCV
2023, 2025
ECCV
2024, 2026
NeurIPS
2023, 2026
ICLR
2024–2027
WACV
2025
AAAI
2024
ACCV
2022, 2024, 2026

Journal reviewing

IEEE TPAMI
2025

Honors & awards

  1. 2026

    Best Patent Award

    Samsung-POSTECH Research Center

  2. 2026

    BK21 Best Paper Award

    POSTECH CSE

  3. 2025

    Outstanding Reviewer

    ICCV

  4. 2023

    BK21 Best Paper Award

    POSTECH CSE