General Information
| Full name | Pyomin Hong |
| pyomindl@g.hongik.ac.kr | |
| Languages | Korean, English |
| Research interests | Generative Model, AI Security & Safety, Trustworthy AI |
Education
-
2025.02 M.S. in Computer Engineering
Hongik University, Seoul, Korea
- Advised by Youn Kyu Lee
-
2023.02 B.E. in Information Security
Seoul Women's University, Seoul, Korea
Research Experience
- 2025.03 - 2026.02
Artificial Intelligence and Information Systems Lab
Visiting Researcher, Prof. Albert No
- Designed a decoding-guided watermarking strategy to secure the provenance of discrete diffusion language models
- 2023.03 - 2025.02
Advanced Intelligent Software Lab
Graduate Research Assistant, Prof. Youn Kyu Lee & Prof. Tae Hyung Kim
- Designed zero-shot inverse imaging with untrained generative priors to reconstruct high-fidelity images from corrupted data
- Developed adaptive orthogonal transforms to augment limited sensor datasets against Gaussian noise
- Assessed adversarial vulnerabilities in biometric systems and developed mitigation strategies to enhance the robustness of face and fingerprint recognition
- Developed detection strategies for deepfakes and fingerprint presentation attacks to ensure cross-dataset reliability in real-world scenarios
- 2022.03 - 2023.02
Advanced Intelligent Software Lab
Undergraduate Research Intern, Prof. Youn Kyu Lee
- Identified vulnerabilities in fingerprint authentication and developed strategies to enhance spoof detection and adversarial robustness
- Aligned and stitched drone thermal images to build a specialized dataset for deep learning–based environmental monitoring
Research Projects
- 2024.04 - 2024.11
AI-based Flood Risk Level Assessment through Video and Image Analysis
Supported by Korea Institute of Science and Technology Information (KISTI)
- Architected a flood risk analysis pipeline integrating YOLO-based object detection and SAM-based segmentation for real-time flood monitoring
- 2022.03 - 2024.12
Development of Integrated Framework for Automated Generation of Neural Network Applications and Optimization of Execution Environment
Supported by Institute for Information communication Technology Planning (IITP)
- Led a team in developing a drag-and-drop-based visual interface for designing and inspecting deep learning architectures
- Designed and implemented neural network architectures, leveraging Docker to ensure reproducible environments
Honors & Awards
- 2025.01
Commendation for Distinguished Achievement
Hongik University
- 2024.10
Best AI Paper Award
IEEE International Conference on ICT Convergence (ICTC)
- 2023.12
AI Utilization Award, Advertising and Promotion Competition (Poster Category)
Gyeonggi Provincial Goverment
- 2022.10
Best Paper Award
IEEE International Conference on ICT Convergence (ICTC)
- 2021.07
Best Project Award, Bahrom Moral Education III (TPBL Course)
Seoul Women’s University
- 2020.08
Excellence Award, Programming GURU1 Hackathon Contest
National Program of Excellence in Software, Seoul Women’s University
Teaching Experience
- Spring 2024
Teaching Assistant, Information Security
Hongik University, Professor: Youn Kyu Lee
- Assisted over 90 Computer Engineering students with hands-on practice covering web hacking and Android vulnerability analysis
- Fall 2024
Teaching Assistant, C-Programming
Hongik University, Professor: Youn Kyu Lee
- Assisted over 30 beginners with hands-on practice, covering basic C programming
Extracurricular Activities
- 2022.09 - 2022.11
Super Hackathon 2022 (NEXON X LIKELION)
NEXON Korea
- Conceptualized and developed a story-driven RPG using Lua, focusing on user progression and engagement
- 2022.09 - 2022.10
BigContest: Field of Data Analysis
Ministry of Science and ICT (MSIT), National Information Society Agency (NIA)
- Analyzed user interaction data to predict loan application outcomes using Random Forest and XGBoost, evaluating feature importance and comparing model performance
- 2022.06 - 2022.08
K-Shield Junior 8th Cohort
Ministry of Science and ICT (MSIT), Korea Internet & Security Agency (KISA)
- Completed an intensive program in cybersecurity fundamentals and incident response, culminating in the development of a Linux-based automated forensics tool
- 2021.07 - 2021.08
Light IT Up Hackathon
Women's University Union IT Hackathon
- Developed an Android application in Java for a designated driver service using Kakao Navigation API for real-time driver tracking and route estimation, selected as a finalist
Technical Skills
| Programming Languages | |
| Python, C/C++, Java, JavaScript |
| Frameworks & Libraries | |
| Django, React |
| Machine Learning | |
| PyTorch, TensorFlow, ONNX (Model Conversion), Hugging Face |
| Infrastructure & Tools | |
| PostgreSQL, Docker, Git, Ubuntu, Figma, Illustrator |