Hello, I am Harim Kim.

I hold a Master’s degree in the Department of Computer Science and Electrical Engineering. My research interest lies in developing innovative deep learning frameworks to address constrained challenges in image and video data analysis. I completed my M.S. in Computer Science at Handong Global University under the supervision of Prof. Charmgil Hong in Handong Artificial Intelligence Lab. (HAIL), and I am currently seeking Ph.D. opportunities in computer science.

RESEARCH INTEREST

• Developing intent-driven deep learning frameworks to address constrained challenges in image and video data analysis
• Applying advanced anomaly detection techniques to medical, industrial, and surveillance domains
• Exploring multimodal data fusion strategies to enhance model performance and enable precise representation learning

EDUCATION

• Master of Computer Science
  • Handong Global University, Pohang, South Korea
  • Graduated: August, 2025
• Bachelor of Computer Science
  • Handong Global University, Pohang, South Korea
  • Graduated: February, 2023

WORK EXPERIENCE

• Graduate Research Assistant
  • Handong Global University, Pohang, South Korea
  • Advisor: Charmgil Hong
  • March 2023 - August 2025 (2.5 years)
• Visiting Researcher
  • University of California, Santa Cruz, CA
  • Advisor: Yuyin Zhou
  • December 2024 - April 2025 (5 month)
• Research Intern
  • Electronics and Telecommunications Research Institute (ETRI), Daejeon, South Korea
  • January 2023 - February 2023 (1.5 month)

REFEREED PUBLICATIONS

• International Venues
  • Harim Kim, Yuhan Wang, Minkyu Ahn, Heeyoul Choi, Yuyin Zhou, Charmgil Hong. Harnessing EHRs for Diffusion-based Anomaly Detection on Chest X-rays. Medical Image Computing and Computer Assisted Intervention (MICCAI). 2025. (Early Accepted) (arxiv)
  • Harim Kim, Chang Ha Lee, Charmgil Hong. Transformer for Point Anomaly Detection. ACM International Conference on Information and Knowledge Management (CIKM). 2024. (paper, github)
  • Jaemin Jung, Soonyong Park, Harim Kim, Chang Ha Lee, Charmgil Hong. Artificial Intelligence-Driven Video Indexing for Rapid Surveillance Footage Summarization and Review. International Joint Conference on Artificial Intelligence (IJCAI) Demo Track. 2024. (paper, github)
  • Harim Kim, Minchae Kim, Taewoo Kim, Kyujin Cho, Charmgil Hong. AREST: Attention-Based Red-Light Violation Detection for Safety Technology. IEEE International Conference on Advanced Video and Signal-Based Surveillance (AVSS). 2024. (paper)
  • Harim Kim, Chang Ha Lee, Charmgil Hong. VATMAN: Video Anomaly Transformer for Monitoring Accidents and Nefariousness. IEEE International Conference on Advanced Video and Signal-Based Surveillance (AVSS). 2024. (paper, github)
  • Harim Kim, Chang Ha Lee, Charmgil Hong. Crime Scene Detection in Surveillance Videos Using Variational AutoEncoder-Based Support Vector Data Description. Asian Conference on Intelligent Information and Database Systems (ACIIDS). 2023. (paper)
• Domestic Venues
  • Harim Kim, Yeseul Kim, Yohan Sohn, Hyo Jung Cho, Soon Sun Kim, Ji Eun Han, Jae Youn Cheong, Charmgil Hong. Prediction of Drug-Induced Liver Injuries Using Transformer. KIISE Transactions on Computing Practices (KTCP). 2024.
  • Harim Kim, Yeseul Kim, Yohan Sohn, Hyo Jung Cho, Soon Sun Kim, Ji Eun Han, Jae Youn Cheong, Charmgil Hong. Prediction of Drug-Induced Liver Injuries Using Transformer. Korea Software Congress (KSC). 2023.
  • Harim Kim, Yuseok Bae, Joong-won Hwang. Distance-Based Confidence Calibration Method Robust to Open Set Classfication. Korea Computer Congress (KCC). 2023.
  • Harim Kim, Charmgil Hong. Outlier Detection-Based Crime Scene Detection in Surveillance Videos Using Variational AutoEncoder. Korea Computer Congress (KCC). 2022.

HONORS

• Best Thesis Award. Handong Global University. August 2025.
• Certificate of Appreciation. Korean Institute of Information Scientists and Engineers (KIISE). July 2025
• Best Paper Award. Handong Global University. November 2024.
• Best Paper Award. 20th IEEE International Conference on Advanced Video and Signal-Based Surveillance. July 2024.
• Best Presentation Award. The Liver Week. June 2024.
• Best Paper Award. Korea Multimedia Society, Spring. May 2024.
• Best Paper Award. KSC (Korean Software Congress). December 2023.
• Best Capstone Project in AI Award. Korea Multimedia Society, Spring. May 2022.
• Best Capstone Design. Handong Global University. June 2022.

PRESENTATIONS

• Transformer for Point Anomaly Detection (Top Conference Session), Korea Computer Congress, South Korea, Jul 2025
• Personalized Prediction of Drug-Induced Liver Injury Using the Transformer Architecture, The Liver Week (TLW), South Korea, Jun 2024
• Red-Light Violation Detection Using Attention Mechanism in Dashcam Footage, Korea Multimedia Society, South Korea, May 2024
• A Method for Lane Recognition and Departure Detection Using Dashcam Videos, Korea Multimedia Society, South Korea, May 2024
• Lenvatinib Efficacy Prediction Using Liver CT Images and a Pre-trained 3D-CNN, Korean Society of Medical Informatics, South Korea, Nov 2022
• Trespass Detection in Surveillance Videos Using VAE- SVDD with 3D-CNN, Korea Multimedia Society, South Korea, May 2022

PARTICIPATED PROJECTS

• Development of Next-Generation Healthcare Machine Learning Technologies Using Multimodal Data
  • Affiliated Institution: Handong Global University, South Korea
  • Collaborating Institution: University of California, Santa Cruz, United States
• Development of High Performance Visual BigData Discovery Platform for Large-Scale Real-time Data Analysis
  • Affiliated Institution: Handong Global University, South Korea / ETRI, South Korea
  • Collaborating Institution: GMDSOFT Inc., South Korea
• Expansion of Connected Car Services through the Development of a Video Analysis-based Safe Driving Evaluation Service
  • Affiliated Institution: Handong Global University, South Korea
  • Collaborating Institution: AiCAR Inc., South Korea
• Development of a Fecal Image-analysis Model for IoT Smart Toilet Services
  • Affiliated Institution: Handong Global University, South Korea
  • Collaborating Institution: HEM Pharma Inc., South Korea
• Development of Drug-induced Liver Injury Prediction Model Utilizing Electronic Medical Records
  • Affiliated Institution: Handong Global University, South Korea
  • Collaborating Institution: Ajou University Hospital, South Korea

TEACHING EXPERIENCE

• Project Advisor, Gyeongbuk Software Meister High School, South Korea
  • Real-time Fire Detection Using AI
• Lecturer, Korea Maritime Meister High School, South Korea
  • One-day Lecture: OpenCV and Artificial Intelligence
• Teaching Assistant, UNESCO UNITWIN, Cambodia
  • Data Science Camp in Cambodia
• Teaching Assistant, Handong Global University, South Korea
  • 2023 / 2024 Machine Learning Camp
  • ITP30010 Database System

VOLUNTEERING

• Student Volunteer, IJCAI (International Joint Conferences on Artificial Intelligence), August 2024

MEDIA COVERAGE

• “Handong Global University and UCSC Kick Off Global Research on Next-Generation Hearthcare AI” - Nate News, December 2024 (link)
• “Handong Global University AI Research Team and AiCAR Win Best Paper Award at International Video Security Conference” - AI Times, August 2024 (link)
• “Handong Global University and GMDSOFT Gain International Recognition for AI Video Processing Technology” - Media Gyeongbuk, August 2024 (link)