💡

Junyeong Song

I'm deeply interested in AI/ML, especially in areas such as OCR, synthetic dataset creation, agentic pipelines, and image generation models.

Beyond implementing models, I care about how technology can create real, practical value. Whenever I learn something new, I naturally start asking, "What's beyond this?" or "How can this be turned into something impactful?" That curiosity consistently drives me to explore further.

What motivates me most is fun and curiosity. I genuinely enjoy the learning process itself, and that sense of excitement pushes me to think deeper and build more meaningful systems.

Junyeong Song

🎓 Education

UNIST

Computer Science and Engineering

2020.03 - Present

Military Service

Under Military Service

2022.07 - 2024.01

Daegu il Science High School

Physics & Computer Science

2017.03 - 2020.02

🔬 Research Interests

Synthetic Data Generation Computer Vision Vision LLM Diffusion Models

📄 Publications

📑

KorMedMCQA

Multi-Choice Question Answering Benchmark for Korean Healthcare Professional Licensing Examinations

arXiv →

💼 Experiences

CJ AI Center

Intern
2025.02 - 2025.09

Developed a Korean Drama script analysis pipeline using a Multi-Agentic pipeline (LangChain).

Embedded AI LAB

Intern
2024.07 - 2024.09

Conducted research on optimization methods in the inference step of diffusion models.


🛠️ Projects


🏆 Awards & Others

🥇

Awards

Graduated 1st and entered UNIST with Unistar Scholarships

👨‍🏫

Teaching

TA in AI Intro and Programming 1