Jinyoung Park

Welcome to my research profile!

I am a Ph.D. candidate in CILab at KAIST EE, advised by Professor Changick Kim. Previously, I was a research intern at Microsoft Research Asia.

My research focuses on multimodal AI systems that perceive, understand, and act in complex real-world environments. My work spans multi-sensor fusion, vision-language and video understanding, and efficient sequence modeling with state space models (SSMs). More recently, I have been developing benchmarks and evaluation methods for multimodal agents that create and edit real-world documents in applications such as Microsoft Word.

If you are interested in collaborating, feel free to reach out :)

news

Apr 29, 2026	One paper got accepted to ICIP 2026! See you in Tampere
Dec 15, 2025	My internship at Microsoft Research Asia has been extended for another 6 months!
Jul 01, 2025	Started a research internship at Microsoft Research Asia!
Dec 09, 2024	One paper got accepted to AAAI!
Jul 01, 2024	Two papers(including VideoMamba) got accepted to ECCV!

selected publications

Towards Efficient Vision State Space Models via Token Merging

Jinyoung Park, Minseok Son, and Changick Kim

ICIP, 2026

Bib PDF

@article{park2025mame,
  author = {Park, Jinyoung and Son, Minseok and Kim, Changick},
  title = {Towards Efficient Vision State Space Models via Token Merging},
  journal = {ICIP},
  year = {2026},
}

Difficulty-aware Balancing Margin Loss for Long-tailed Recognition

Minseok Son, Inyong Koo, Jinyoung Park, and Changick Kim

AAAI, 2025

Bib PDF

@article{son2024difficulty,
  title = {Difficulty-aware Balancing Margin Loss for Long-tailed Recognition},
  author = {Son, Minseok and Koo, Inyong and Park, Jinyoung and Kim, Changick},
  journal = {AAAI},
  year = {2025},
}

VideoMamba: Spatio-Temporal Selective State Space Model

Jinyoung Park, Hee-Seon Kim, Kangwook Ko, Minbeom Kim, and 1 more author

ECCV, 2024

Bib PDF

@article{park2024videomamba,
  author = {Park, Jinyoung and Kim, Hee-Seon and Ko, Kangwook and Kim, Minbeom and Kim, Changick},
  title = {VideoMamba: Spatio-Temporal Selective State Space Model},
  journal = {ECCV},
  year = {2024},
}

Flow-Assisted Motion Learning Network for Weakly-Supervised Group Activity Recognition

Muhammad Adi Nugroho, Sangmin Woo, Sumin Lee, Jinyoung Park, and 3 more authors

ECCV, 2024

Bib PDF

@article{nugroho2024flamingnet,
  author = {Nugroho, Muhammad Adi and Woo, Sangmin and Lee, Sumin and Park, Jinyoung and Wang, Yooseung and Kim, Donguk and Kim, Changick},
  title = {Flow-Assisted Motion Learning Network for Weakly-Supervised Group Activity Recognition},
  journal = {ECCV},
  year = {2024},
}

Sketch-based Video Object Localization

Sangmin Woo, So-Yeong Jeon, Jinyoung Park, Minji Son, and 2 more authors

WACV, 2024

Bib PDF

@article{woo2023svol,
  author = {Woo, Sangmin and Jeon, So-Yeong and Park, Jinyoung and Son, Minji and Lee, Sumin and Kim, Changick},
  title = {Sketch-based Video Object Localization},
  journal = {WACV},
  year = {2024},
}

Anchoring Vision and Language Knowledge for Weakly Supervised Group Activity Recognition

Muhammad Adi Nugroho, Jinyoung Park, Donguk Kim, and Changick Kim

VCIP, 2024

Bib PDF

@article{nugroho2024anvil,
  author = {Nugroho, Muhammad Adi and Park, Jinyoung and Kim, Donguk and Kim, Changick},
  title = {Anchoring Vision and Language Knowledge for Weakly Supervised Group Activity Recognition},
  journal = {VCIP},
  year = {2024},
}

Multi-modal Social Group Activity Recognition in Panoramic Scene

Donguk Kim, Sumin Lee, Sangmin Woo, Jinyoung Park, and 2 more authors

VCIP, 2023

Bib PDF

@article{kim2023multi,
  author = {Kim, Donguk and Lee, Sumin and Woo, Sangmin and Park, Jinyoung and Nugroho, Muhammad Adi and Kim, Changick},
  title = {Multi-modal Social Group Activity Recognition in Panoramic Scene},
  journal = {VCIP},
  year = {2023},
}

Rainunet for super-resolution rain movie prediction under spatio-temporal shifts

Jinyoung Park, Minseok Son, Seungju Cho, Inyoung Lee, and 1 more author

NeurIPS Weather4Cast Challenge, 2022

Bib PDF

@article{park2022rainunet,
  author = {Park, Jinyoung and Son, Minseok and Cho, Seungju and Lee, Inyoung and Kim, Changick},
  title = {Rainunet for super-resolution rain movie prediction under spatio-temporal shifts},
  journal = {NeurIPS Weather4Cast Challenge},
  year = {2022},
}

Nowformer: A Locally Enhanced Temporal Learner for Precipitation Nowcasting

Jinyoung Park, Inyoung Lee, Minseok Son, Seungju Cho, and 1 more author

NeurIPS Tackling Climate Change with Machine Learning Workshop, 2022

Bib PDF

@article{park2022nowformer,
  author = {Park, Jinyoung and Lee, Inyoung and Son, Minseok and Cho, Seungju and Kim, Changick},
  title = {Nowformer: A Locally Enhanced Temporal Learner for Precipitation Nowcasting},
  journal = {NeurIPS Tackling Climate Change with Machine Learning Workshop},
  year = {2022},
}

Dat: Domain adaptive transformer for domain adaptive semantic segmentation

Jinyoung Park, Minseok Son, Sumin Lee, and Changick Kim

ICIP, 2022

Bib PDF

@article{park2022dat,
  author = {Park, Jinyoung and Son, Minseok and Lee, Sumin and Kim, Changick},
  title = {Dat: Domain adaptive transformer for domain adaptive semantic segmentation},
  journal = {ICIP},
  year = {2022},
}

Explore-and-match: Bridging proposal-based and proposal-free with transformer for sentence grounding in videos

Sangmin Woo, Jinyoung Park, Inyong Koo, Sumin Lee, and 2 more authors

arXiv, 2022

Bib PDF

@article{woo2022explore,
  author = {Woo, Sangmin and Park, Jinyoung and Koo, Inyong and Lee, Sumin and Jeong, Minki and Kim, Changick},
  title = {Explore-and-match: Bridging proposal-based and proposal-free with transformer for sentence grounding in videos},
  journal = {arXiv},
  year = {2022},
}