I am a final-year undergraduate at Sangmyung University, double-majoring in Electronic Engineering and Human-Centered AI Engineering. My research focuses on machine learning for computer vision, the bio domain, and on-device inference.
For my master's degree, I aim to study model compression and optimization, how training data affects model performance, and reinforcement learning.
- [1st author] The choice of species representation in multi-species aquatic acute-toxicity prediction is constrained by data availability rather than by predictive accuracy — Journal of Cheminformatics (SCIE, Q1), under review
- [Co-first author] Dissecting Tissue-Context Effects in Graph Neural Networks for Rat Kp Prediction: Additive Shifts, Fusion Position, and Extrapolation — Journal of Pharmacokinetics and Pharmacodynamics (SCIE, Q2), under review
- [Co-author] Implementation and Processing-Time Analysis of an NPU-Based Real-Time Iris Recognition System — IEIE Summer Conference 2025
| Project | Description |
|---|---|
| Bifusion | Normal/Anomaly classifier trainer with diffusion-based augmentation (Medfusion) and few-shot learning (ProtoNet) |
| Mouse Filler Volume Prediction | 3D reconstruction from 2D video with Instant-NGP — Gachon Univ. |
| Tissue-specific Kp Prediction | Chemprop-based GNN, lead researcher — KRIBB |
| NPU Algorithm for Toxicity Prediction | MLP built only from NPU-supported operators on Kneron KL630 — KRIBB |
| DoorBox | YOLOv5s face detection on the Kneron KL630 NPU; emotion and gender/age classifiers (MobileNetV3-Small via knowledge distillation) on the CPU |
See all projects on my portfolio.
Languages
ML & Data
Computer Vision & Generative Models
Deployment & Hardware
OS & Tools