Jin-Hwi Park (박진휘)
Ph.D. Candidate jinhwipark [at] gm.gist.ac.kr
I received the BS dual degree in Mathematics and Convergence Software from Sogang University in 2020. I am a Ph.D. candidate at Gwangju Institute of Science and Technology, under the supervision of Prof. Hae-Gon Jeon. My research focuses on computer vision and the interdisciplinary application of AI, integrating insights from mathematics, engineering, and diverse domains to address real-world challenges.
NEWS
09/2024: 1 paper accepted to NeurIPS 2024
07/2024: I attend ICVSS 2024 at Sicily, Italy
04/2024: 1 paper accepted to TPAMI (Impact Factor: 20.8)
02/2024: 1 paper accepted to CVPR 2024
01/2024: 1 paper accepted to ICLR 2024
09/2023: 1 paper accepted to Advanced Science (Impact Factor: 14.3)
08/2023: 1 paper accepted to Nano Today (Impact Factor: 13.2)
04/2023: 1 paper accepted to ICML 2023
09/2022: I started my internship at Bot Intelligence Group, The Robotics Institute, CMU.
(Contributed as a first author)
RESEARCH INTERESTS
Computer Vision: Low-level Vision[4, 6, 8, 10], 3D Vision[4, 7, 8, 10]
Robotics: Sensor Fusion[4, 7, 10], Environmental Sensing[1, 7, 9], Trajectory Prediction[2, 3]
Interdisciplinarity: Mathematics[2, 4, 7], Optics[5, 6], Social Science[1, 9]
EDUCATION
Gwangju Institute of Science and Technology (GIST), Gwangju, Korea
Ph.D., Artificial Intelligence (2020 - present)
Advisor: Hae-Gon JeonSogang University, Seoul, Korea
B.S., Mathematics & B.S.E., Convergence Software (2015 - 2020)
PUBLICATIONS
[10] A Simple yet Universal Framework for Depth Completion
Jin-Hwi Park and Hae-Gon Jeon
The Thirty-eighth Annual Conference on Neural Information Processing Systems (NeurIPS), 2024
[Paper] / [Source Code]
[9] What Makes Deviant Places?
Jin-Hwi Park*, Young-Jae Park*, Ilyung Cheong, Junoh Lee, Young Eun Huh, and Hae-Gon Jeon
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) (IF: 20.8), 2024
[Paper] / [Source Code]
[8] Depth Prompting for Sensor-agnostic Depth Estimation
Jin-Hwi Park, Chanhwi Jeong, Junoh Lee and Hae-Gon Jeon
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Jun 2024
[Paper] / [Source Code]
[7] Geometry-Aware Projective Mapping for Unbounded Neural Radiance Fields
Junoh Lee, Hyunjun Jung, Jin-Hwi Park, Inhwan Bae and Hae-Gon Jeon
The Twelfth International Conference on Learning Representations (ICLR), 2024
[Paper] / [Source Code]
[6] DeepGT: Deep Learning-based Quantification of Nanosized Bioparticles in Brightfield Micrographs of Gires-Tournois Biosensor
Jiwon Kang*, Young Jin Yoo*, Jin-Hwi Park*, Joo Hwan Ko, Seungtaek Kim, Stefan G. Stanciu, Harald A. Stenmark, Jin Ah Lee, Abdullah Al Mahmud, Hae-Gon Jeon and Young Min Song
Nano Today (IF: 13.2), 2023
[Paper] / [Source Code]
[5] Full-control and Switching of Optical Fano Resonance by Continuum State Engineering
Joo Hwan Ko*, Jin-Hwi Park*, Young Jin Yoo*, Sehui Chang, Jiwon Kang, Aiguo Wu, Fang Yang, Sejeong Kim, Hae-Gon Jeon and Young Min Song
Advanced Science (IF: 14.3), 2023
[Paper] / [Source Code]
[4] Learning Affinity with Hyperbolic Representation for Spatial Propagation
Jin-Hwi Park, Jaesung Choe, Inhwan Bae and Hae-Gon Jeon
The Fortieth International Conference on Machine Learning (ICML), July 2023
[Paper] / [Source Code]
[3] Learning Pedestrian Group Representations for Multi-modal Trajectory Prediction
Inhwan Bae, Jin-Hwi Park and Hae-Gon Jeon
European Conference on Computer Vision (ECCV), Oct 2022
[Paper] / [Source Code]
[2] Non-Probability Sampling Network for Stochastic Human Trajectory Prediction
Inhwan Bae, Jin-Hwi Park and Hae-Gon Jeon
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Jun 2022
[Paper] / [Source Code]
[1] DevianceNet: Learning to Predict Deviance from A Large-scale Geo-tagged Dataset
Jin-Hwi Park*, Young-Jae Park*, Junoh Lee and Hae-Gon Jeon
The Thirty-Sixth AAAI Conference on Artificial Intelligence (AAAI), Feb 2022
[Paper] / [Source Code & Dataset]
HONORS & AWARDS
Top Scholar, Student Research Achievement Scholarship, GIST, Dec 2024. [Certificate]
Winner, Qualcomm Innovation Fellowship Korea, Qualcomm Technologies, Inc., Dec 2024. [News] [Certificate]
Top Reviewers of NeurIPS 2024, Vancouver, Canada, Dec 2024. [link]
11th Sunseom Scholarship, Development Foundation, GIST, Nov 2024. [link] [Certificate]
International Computer Vision Summer School, Sicily, Italy, Jul 2024. [Certificate of Attendance]
International Computer Vision Summer School Examination, Sicily, Italy, Jul 2024. [Examination Certificate]
RA Student Research Achievement Scholarship, AI Graduate School, GIST, Dec 2023.
RA Student Research Achievement Scholarship, AI Graduate School, GIST, Dec 2022.
Finalist, 2022 Qualcomm Innovation Fellowship Korea, Qualcomm Technologies, Inc., Nov 2022.
Overseas Internship Program Scholarship Funding, AI Graduate School, GIST, Sep 2022.
PATENT
Model learning method and system capable of sensor-agnostic depth map inference through depth prompting, and depth map inference method and system using the same (US).
US Patent App. 19/002,363, Dec 2024Method and system for implementing radiance field using adaptive mapping function (US).
US Patent App. 19/002,445, Dec 2024Depth map estimation method and system using a universal framework (KR).
KR Patent App. 10-2024-0177675, Dec 2024Radiance field implementation method and system using adaptive mapping function. (KR)
KR Patent App. 10-2024-0044389, Apr 2024Learning method and system capable of sensor-agnostic depth map estimation through depth prompting. (KR)
KR Patent App. 10-2024-0041415, Mar 2024Parameter determination method for porous fano resonant structure. (KR)
KR Patent App. 10-2023-0085357, Jul 2023Optical microscopy image enhancement method using AI model. (KR)
KR Patent App. 10-2023-0094453, Jun 2023Pedestrian trajectory prediction method using non‑probability sampling. (US) [Google Patents]
US Patent App. 18/174,716, Feb 2023Image convolutional method in hyperbolic space. (US) [Google Patents]
US Patent App. 17/813,963, Jul 2022Image convolutional method in hyperbolic space. (KR) [Certificate of Patent]
KR Patent No. 10-2388760, Apr 2022Pedestrian trajectory prediction method using grouping pedestrians. (KR) [Certificate of Patent]
KR Patent App. 10-0052202, Apr 2022Pedestrian trajectory prediction method using non‑probability sampling. (KR)
KR Patent App. 10-2022-0032099, Mar 2022Safety information provision apparatus. (KR) [Certificate of Patent]
KR Patent No. 10-2518331, Mar 2021Deviance recognition apparatus. (KR)
KR Patent App. 10-2020-0157403, Nov 2020
RESEARCH EXPERIENCE
Carnegie Mellon University, Pittsburgh, USA
Visiting Researcher, Robot Intelligence Group (Sep. 2022 - Feb. 2023)
Advisor: Jean OhSogang University, Seoul, Korea
Undergraduate Researcher, Bigdata Processing & DB Lab (Mar. 2019 - Dec. 2019)
Advisor: Sungwon JungPurdue University, West Lafayette, USA
Visiting Student, K-SW Square (Jul. 2018 – Aug. 2018)
Advisor: Eric Matson
Contact
Tel: +82 - 62 - 715 - 6375
Email: jinhwipark (at) gist.ac.kr
Address:
GIST, Electrical Engineering and Computer Science Building C, Room 412
123, Cheomdangwagi-ro, Buk-gu, Gwangju, Republic of Korea