Hello, I am Gilhan Park, a Ph.D. student in Artificial Intelligence at Sungkyunkwan University, South Korea.
I am a member of the Visual Computing LAB (VCLAB) advised by Prof. Jae-Pil Heo.
My research interests lie in Computer Vision, with a particular focus on Segmentation (2D-to-3D Lifting) and Continual Learning.
Specifically, I am working on expanding 2D visual representations into 3D structural understanding. In addition, I study Continual Learning methods that enable models to flexibly update and adapt to new tasks over time.
If you are interested in research discussions or potential collaborations, please feel free to reach out to me via email.
Research Interests
2D Segmentation
3D Segmentation (2D-to-3D Lifting)
Continual Learning
Video Understanding
News
Sep. 2025
🔥 Our paper 'Mitigating Semantic Collapse in Partially Relevant Video Retrieval' was accepted to the 39th Conference on Neural Information Processing Systems (NeurIPS 2025).
Mar. 2025
Continuing my Ph.D. program in Artificial Intelligence at Sungkyunkwan University, conducting research as a member of the Visual Computing LAB (VCLAB).
Mar. 2025
Receiving continuous support as a Ph.D. scholarship recipient of the Hyundai Motor Chung Mong-Koo Foundation.
Selected Projects
Click to see details
Sep. 2025
Mitigating Semantic Collapse in Partially Relevant Video Retrieval
Addressing semantic collapse in text and video embeddings for Partially Relevant Video Retrieval
Jul. 2024
Mitigating Background Shift in Class-Incremental Semantic Segmentation
Addressing Background Shift in Class-Incremental Semantic Segmentation
Educations
Mar. 2025 - Present
Sungkyunkwan University, Suwon, South Korea
Ph.D. in Artificial IntelligenceMar. 2023 - Feb. 2025
Sungkyunkwan University, Suwon, South Korea
Master of Engineering in Artificial IntelligenceGPA: 3.94/4.5
Mar. 2017 - Feb. 2023
Sungkyunkwan University, Seoul, South Korea
Bachelor of Science in Computer Science and EducationGPA: 3.81/4.5
Mar. 2013 - Feb. 2016
Publications
Mitigating Semantic Collapse in Partially Relevant Video Retrieval
WonJun Moon*, MinSeok Jung*, Gilhan Park, Tae-Young Kim, Cheol-Ho Cho, Woojin Jun, and Jae-Pil HeoProceedings of the 39th Conference on Neural Information Processing Systems (NeurIPS 2025)
Mitigating Background Shift in Class-Incremental Semantic Segmentation
Gilhan Park, WonJun Moon, SuBeen Lee, Tae-Young Kim, and Jae-Pil HeoProceedings of Conference on 2024 IEEE European Conference on Computer Vision (ECCV), Milano, Italy
Improving Classification Results on Medical Dataset Augmentation using GAN
Gilhan Park, Jaehwan Lee, Ikgyu Lee, Jisoo Kim, and Jae-Pil HeoProceedings of Korea Software Congress, 2022, (KSC 2022)
Edge Computing Based Cloud Parking Management System
Gilhan Park, Heedong Kim, Serin Yoon, Jaehwan Lee, and Soyeon JeongProceedings of the Korea Information Processing Society Conference. KIPS, 2021
Experiences
Jul. 2022 - Present
Sungkyunkwan University, Suwon, South Korea
Researcher, Visual Computing LabConducting research on 2D Segmentation and 3D Segmentation (2D-to-3D Lifting). In parallel, working on Continual Learning and Incremental Learning for building adaptive vision systems.
Oct. 2024 - Nov. 2024
Teaching Assistant for LG Electronics Manufacturing AI Leadership Program
Teaching AssistantAssisted in the LG Electronics Manufacturing AI Leadership Program by supporting project-based learning and providing guidance on AI methodologies.
Dec. 2021 - Jun. 2022
Sungkyunkwan University, Suwon, South Korea
Undergraduate Research Student, CSI LabConducted research on improving the performance of Semantic Segmentation in remote sensing datasets.
Jul. 2021 - Dec. 2021
Smilegate Future Lab, Pangyo, South Korea
AI Project Contributor, SGM AI ProjectDeveloped real-time clothing pattern generation and style transfer under the guidance of the Smilegate AI team.
Honors & Awards
May. 2023 - Present
Hyundai Motor Company Chung Mong-koo Scholarship
Selected as a recipient of the Hyundai Motor Chung Mong-koo Foundation Scholarship
Nov. 2024
Qualcomm Innovation Fellowship South Korea 2024
FinalistSelected as a finalist for the Qualcomm Innovation Fellowship South Korea 2024 with the paper 'QIFK2024_69: Mitigating Background Shift in Class-Incremental Semantic Segmentation.'
Sep. 2022
2nd SKKU AI-SW Education Hackathon, Gold Prize
CodeWithMe: Extracted code images during programming lectures, utilizing a model for Document Understanding, typo correction, and code comprehension
Dec. 2021
Outstanding Paper Award at ACK 2021, Korea Information Processing Society, Outstanding Paper Award
Received Outstanding Paper Award at the ICT Conference of Korea Information Processing Society (ACK 2021) with the paper 'Edge Computing-Based Cloud Parking Management System'
Dec. 2021
1st SKKU SW-AI Education Capstone Design Competition, Grand Prize
Implemented the game 'Red Light, Green Light' from Netflix's 'Squid Game' using Jetson Nano
Dec. 2021
2021 Hanium Competition, Honorable Mention
Developed a Korean license plate recognition model using LPRnet and parking automation program
Oct. 2021
1st SKKU AI-SW Education Hackathon, Silver Prize
Focus Maker: A service to measure online learners' concentration index and provide self-feedback
Dec. 2021
Completion of Smilegate AI Membership
Participated in Smilegate AI Membership
Aug. 2021
2021 AI Online Competition, 3rd Place
Won 3rd place in the AI Online Competition hosted by the Ministry of Science and ICT with a model detecting driver negligence behavior for accident prevention
Copyright © 2023 Jiwon Jason Choi. All Rights Reserved.