Sogang University, Seoul, Korea
Mar. 2026 – PresentM.S. Student, Department of Artificial Intelligence
Intelligent Information Processing Lab (IIP) · Advisor: Prof. Hyung-Min Park
Academic & professional background
Research in speech recognition and audio-language models, alongside experience in applied machine learning and embedded systems.
M.S. Student, Department of Artificial Intelligence
Intelligent Information Processing Lab (IIP) · Advisor: Prof. Hyung-Min Park
B.S. in Electronic Engineering
GPA 3.98 / 4.5
Intelligent Information Processing Lab (IIP), Sogang University
Intelligent Information Processing Lab (IIP), Sogang University · Presentation (PDF)
Machine Intelligence Lab, Inha University
Intelligence Embedding System Lab (IESL), Inha University
Jongha Kim, Leehyeon Song, Hyung-Min Park. “Task-Leaf Routed MiMo-Audio for DCASE 2026 Task 5.” DCASE 2026 Challenge, technical report, Jun. 2026. Official evaluation accuracy: 49.13% (Kim_SGU_task5_4). Report (PDF) · Challenge results
Jang Ji-hye, Lee Young-jun, Heo Ji-won, Kim Jong-ha. “Development of an ROS-based Environmental Perception and Decision-making System for Indoor Autonomous Mobile Robots.” Korean Society of Automotive Engineers (KSAE), Jeju, Korea, Oct. 2022. (Poster)
R&D Department, STS Engineering
ECE Capstone Design, Inha University · GitHub
Exploring Korean–English speech-to-speech translation through discrete speech units, without an intermediate text representation at inference time.
Generated English speech scored BLEU 42.8 / COMET 0.1009 in the project evaluation. Output was intelligible but had unstable pitch and background noise.
Machine Intelligence Lab, Inha University
Explored controllable video generation with scene graphs as structured conditions for a diffusion model.
Implemented the full pipeline. Generation quality remained limited by scarce scene-graph–video paired data (Action Genome) and weak graph–video alignment.
2024 Inha Artificial Intelligence Challenge (Dacon) · GitHub
Explored question answering over Korean economic articles through language-model fine-tuning.
LoRA adaptation degraded the already well-aligned base model in project experiments; the approach shifted to minimal fine-tuning to preserve base behavior.
6th AI SPARK Challenge · GitHub
Built a satellite-image segmentation system for wildfire detection using an ensemble of TransUNet and Attention U-Net.
Sogang University, Dept. of Electronic Engineering
Semiconductor Elite Project (EEE3150)
Sogang University, Dept. of Electronic Engineering
Advanced Engineering Mathematics I (EEE2103)
Inha University, Dept. of Electronic Engineering
Deep Learning · Introduction to Machine Learning