I am a postdoctoral researcher in Department of Computer Science and Engineering at Pohang University of Science and Technology (POSTECH) and a member of the Computer Vision Lab.. I completed my Ph.D. in 2024, under the supervision of Minsu Cho at POSTECH, Pohang, Korea. My research interest is computer vision and deep learning, especially scene graph generation and neuro-symbolic learning.

News

• [Sep. 2024] I have started a new role as a Postdoctoral reseracher at POTSECH, working with Prof. Minsu Cho.
• [Aug. 2024] I am glad to announce that I successfully defended my Ph.D. dissertation, “Scene Understaning with Contextual Reasoning via Message Passing.”
• [Jun. 2024] A paper about burst-image enhencement is accepted to NTIRE workshop at CVPR 2024.
• [Sep. 2023] A paper about neural symbolic methods for temporal action segmentation is accepted to NeurIPS 2023.
• [Feb. 2023] Two papers are accepted to CVPR 2023.
• [Mar. 2022] A paper about few-shot metric learning is accepted to ACCV 2022.
• [Sep. 2018] A paper about semantic correspondence is accepted to ECCV 2018.

Working Experience

• Postdoctoral researcher at Pohang University on Science and Technology, Pohang, Republic of Korea
  Department of Computer Science and Engineering
  working with Prof. Minsu Cho
  Research focus: scene graph generation, neuro-symbolic learning

Publication

• Scene Understanding with Contextual Reasoning via Message Passing
  Deunsol Jung, Ph.D. dissertation, Aug. 2024
  [paper]
• Burst Image Super-Resolution with Base Frame Selection
  Sanghyun Kim*, Min Jung Lee*, Woohyeok Kim, Deunsol Jung, Jaesung Rim, Sunghyun Cho, Minsu Cho, NTIRE Workshop@CVPR 2024 (* Equal contributions)
  [paper] [project page]
• Activity Grammars for Temporal Action Segmentation
  Dayoung Gong*, Joonseok Lee*, Deunsol Jung, Suha Kwak, Minsu Cho, NeurIPS 2023 (* Equal contributions)
  [paper] [project page] [github]
• Devil’s on the Edges: Selective Quad Attention for Scene Graph Generation
  Deunsol Jung, Sanghyun Kim, Won Hwa Kim, Minsu Cho, CVPR 2023
  [paper] [project page] [github]
• Relational Context Learning for Human-Object Interaction Detection
  Sanghyun Kim, Deunsol Jung, Minsu Cho, CVPR 2023
  [paper] [project page] [github]
• Few-shot Metric Learning: Online Adaptation of Embedding for Retrieval
  Deunsol Jung, Dahyun Kang, Suha Kwak, Minsu Cho, ACCV 2022
  [paper] [github]
• Attentive Semantic Alignment with Offset-Aware Correlation Kernels
  Paul Hongsuck Seo, Jongmin Lee, Deunsol Jung, Bohyung Han, Minsu Cho, ECCV 2018
  [paper] [project page] [github]

Education

• B.S. in Computer Science and Engineering
  POSTECH, Pohang, Republic of Korea
  Mar. 2011 - Aug. 2016, GPA: 3.36/4.30
• Ph.D. (Unified master’s and doctor’s course) in Computer Vision Lab.
  Department of Computer Science and Engineering, POSTECH, Pohang, Republic of Korea
  Sep. 2016 - Aug. 2024, GPA: 3.59/4.30

Awards

• Best Presentation Awards @ IPIU 2019 for outstanding poster presentation.

Teaching & Advising

• Teaching assistant (TA) for undergraduate course
  • Object-Oriented Programming (OOP), Mar. 2017 - Aug. 2017
  • Automata and Formal Language, Sep. 2017 - Dec. 2017
  • Automata and Formal Language, Sep. 2018 - Dec. 2018
• TA for Industry Oriented Course
  • AI expert training course for POSCO, Mar. 2017 - Oct. 2017
    • Introduction to the Python Programming
    • Computer Vision
  • Machine learning expert training course for Samsung DS, May. 2017 - Jun. 2017
    • Introduction to the Python Programming
    • Computer Vision
  • AI programming for the job seeker, Sep. 2017
    • Introduction to the Python Programming
  • AI industry professional training for Gyeongsangbuk-Do, Dec. 2017
    • Computer Vision
  • AI expert training course for POSCO, Jun. 2018 - Jul. 2018
    • Computer Vision
  • AI expert training course for POSCO, Jul. 2019 - Aug. 2019
    • Computer Vision
  • AI expert training course for POSCO, Mar. 2022 - Mar. 2022
    • Introduction to the Python Programming