CV
Academic background, experience, publications, and projects.
Contact Information
| Name | Cho Seok-lea |
| cho020218@gmail.com |
Experience
-
2025.12 - present Hansung University, Seoul
Undergraduate Researcher
Prof. Jinwoo Gong Research Group
- Researching trustworthy federated and distributed AI, including privacy leakage, poisoning, adversarial manipulation, and robust defenses.
- Designing learning frameworks that combine federated unlearning with Bayesian uncertainty estimation to improve reliability, accountability, and security.
- Improving the scalability and efficiency of federated learning through resource-aware sparsification without sharing client data.
-
2026.04 - 2027.02 Korea
Research Intern
High-Performance Computing Lab, Chung-Ang University
-
2024.01 - 2024.03 Participant
LG Aimers 4th
- Completed AI education and project-based learning program.
-
2023.10 - 2023.11 Participant
BoostCourse Coaching Study 13th
- Completed coaching study activities through BoostCourse.
Summary
Education
-
2021 - 2027 Seoul, South Korea
B.S.
Hansung University
IT Convergence Engineering, Intelligent System and IoT Track
Publications
-
2025 Evaluating Generalization and Data Leakage in Large Language Models via Data Perturbation
IEIE Conference, 2025
Perturbation-based robustness test for checking data leakage and generalization in large language model evaluation.
Projects
-
VideoQna
Local video search app that lets users ask questions and retrieve matching scenes, subtitles, summaries, and keywords from indexed video content.
- Video question answering
- Scene and subtitle retrieval
-
Pocket-Coach
IoT capstone app for soccer video analysis with player/ball tracking, highlight feedback, USB MIDI servo control, and ExecuTorch NPU on-device detection.
- Soccer video analysis
- ExecuTorch NPU inference prototype
-
YOLO in iOS with NPU
Swift iOS app for YOLO detection with ExecuTorch inference, live camera mode, still-image tests, and bbox overlay.
- On-device object detection
- iOS live camera inference
Awards
-
Hansung University, Internet of Things Track Capstone Design Excellence Award (2nd place)
-
Hansung University, Intelligence System Track Capstone Design Excellence Award (2nd place)
Skills
AI and Edge Systems: PyTorch, CUDA, ExecuTorch, iOS, Edge AI, Federated Learning