Jungin Park

Assistant Professor, PhD

DGIST, EECS

Daegu, Korea

I am Jungin Park, an assistant professor in the Department of Electrical Engineering and Computer Science at DGIST. Before joining DGIST, I was a research professor from April 2025 to September 2026 and a postdoctoral researcher from September 2024 to March 2025 at Yonsei University, working with Prof. Kwanghoon Sohn. My research interests include computer vision, video understanding, multimodal learning, and embodied AI.

I received my Ph.D. degree from the School of Electrical and Electronic Engineering at Yonsei University in 2024, advised by Prof. Kwanghoon Sohn.

news

Oct 06, 2026	DYNAMIC LAB @ DGIST is now open! 👏
Oct 06, 2026	Our paper on vision-language-action models was accepted to NeurIPS 2026.
May 01, 2026	Our two papers were accepted to ICML 2026.
Nov 18, 2025	Our paper on geospatial domain adaptation was accepted to GRSL.
Sep 28, 2025	Our paper on language-guided video summarization was accepted to IJCV.

selected publications

Embodied AI

Spatial Representation Distillation and Knowledge Routing for Vision-Language-Action Models

Jungin Park^*, Chaoran Zhu^*, Changjae Oh, and 1 more author

In Conference on Neural Information Processing Systems (NeurIPS), 2026

@inproceedings{park2026spatial,
  title = {Spatial Representation Distillation and Knowledge Routing for Vision-Language-Action Models},
  author = {Park, Jungin and Zhu, Chaoran and Oh, Changjae and Sohn, Kwanghoon},
  booktitle = {Conference on Neural Information Processing Systems (NeurIPS)},
  year = {2026},
  publisher = {NeurIPS},
  selected = true
}

Omnimodal

V-LynX: Video interface alignment for video+x llms

Jungin Park, Jiyoung Lee^†, and Kwanghoon Sohn^†

In Internaional Conference on Machine Learning (ICML), 2026

@inproceedings{park2026vlynx,
  title = {V-LynX: Video interface alignment for video+x llms},
  author = {Park, Jungin and Lee, Jiyoung and Sohn, Kwanghoon},
  booktitle = {Internaional Conference on Machine Learning (ICML)},
  year = {2026},
  publisher = {ICML},
  selected = true
}

Efficient ML

Saliency-aware model merging

Jungin Park, Jiyoung Lee^†, and Kwanghoon Sohn^†

In Internaional Conference on Machine Learning (ICML), 2026

@inproceedings{park2026saliency,
  title = {Saliency-aware model merging},
  author = {Park, Jungin and Lee, Jiyoung and Sohn, Kwanghoon},
  booktitle = {Internaional Conference on Machine Learning (ICML)},
  year = {2026},
  publisher = {ICML},
  selected = true
}

Video-language

Language-guided recursive spatiotemporal graph modeling for video summarization

Jungin Park, Jiyoung Lee^†, and Kwanghoon Sohn^†

International Journal of Computer Vision (IJCV), 2025

@article{park2025videograph,
  title = {Language-guided recursive spatiotemporal graph modeling for video summarization},
  author = {Park, Jungin and Lee, Jiyoung and Sohn, Kwanghoon},
  journal = {International Journal of Computer Vision (IJCV)},
  year = {2025},
  publisher = {Springer},
  selected = true
}

Video understanding

Bootstrap your own views: Masked ego-exo modeling for fine-grained view-invariant video representations

Jungin Park, Jiyoung Lee^†, and Kwanghoon Sohn^†

In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025

@inproceedings{park2025bootstrap,
  title = {Bootstrap your own views: Masked ego-exo modeling for fine-grained view-invariant video representations},
  author = {Park, Jungin and Lee, Jiyoung and Sohn, Kwanghoon},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year = {2025},
  publisher = {IEEE/CVF},
  selected = true
}

Vision-language

Bridging vision and language spaces with assignment prediction

Jungin Park, Jiyoung Lee^†, and Kwanghoon Sohn^†

In International Conference on Learning Representations (ICLR), 2024

@inproceedings{park2024briding,
  title = {Bridging vision and language spaces with assignment prediction},
  author = {Park, Jungin and Lee, Jiyoung and Sohn, Kwanghoon},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2024},
  selected = true
}

Efficient ML

Dual-path adaptation from image to video transformers

Jungin Park^*, Jiyoung Lee^*, and Kwanghoon Sohn

In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023

@inproceedings{park2023dual,
  title = {Dual-path adaptation from image to video transformers},
  author = {Park, Jungin and Lee, Jiyoung and Sohn, Kwanghoon},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  pages = {2203--2213},
  year = {2023},
  publisher = {IEEE/CVF},
  selected = true
}

Video understanding

Probabilistic representations for video contrastive learning

Jungin Park, Jiyoung Lee, Ig-Jae Kim, and 1 more author

In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022

@inproceedings{park2022probabilistic,
  title = {Probabilistic representations for video contrastive learning},
  author = {Park, Jungin and Lee, Jiyoung and Kim, Ig-Jae and Sohn, Kwanghoon},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  pages = {14711--14721},
  year = {2022},
  publisher = {IEEE/CVF},
  selected = true
}

Video-language

Bridge to answer: Structure-aware graph interaction network for video question answering

Jungin Park, Jiyoung Lee, and Kwanghoon Sohn

In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021

@inproceedings{park2021bridge,
  title = {Bridge to answer: Structure-aware graph interaction network for video question answering},
  author = {Park, Jungin and Lee, Jiyoung and Sohn, Kwanghoon},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  pages = {15526--15535},
  year = {2021},
  publisher = {IEEE/CVF},
  selected = true
}

Video understanding

Sumgraph: Video summarization via recursive graph modeling

Jungin Park^*, Jiyoung Lee^*, Ig-Jae Kim, and 1 more author

In European Conference on Computer Vision (ECCV), 2020

@inproceedings{park2020sumgraph,
  title = {Sumgraph: Video summarization via recursive graph modeling},
  author = {Park, Jungin and Lee, Jiyoung and Kim, Ig-Jae and Sohn, Kwanghoon},
  booktitle = {European Conference on Computer Vision (ECCV)},
  pages = {647--663},
  year = {2020},
  publisher = {Springer},
  selected = true
}