Tongmin Kim

I'm an undergraduate student at Korea University, majoring in both Health & Environmental Convergence Science and Artificial Intelligence. I am currently working as an undergraduate researcher at CVLAB, KAIST AI, advised by Prof. Seungryong Kim.

I am mainly interested in the problem of 3D/4D Reconstruction & Understanding. Recently, I have been studying world models — exploring how agents can learn internal representations that predict and reason about the dynamics of the physical world.

Email · CV · Google Scholar · GitHub · LinkedIn

Publications

^† denotes corresponding author.

CORAL: Correspondence Alignment for Improved Virtual Try-On

Jiyoung Kim, Youngjin Shin, Siyoon Jin, Dahyun Chung, Jisu Nam, Tongmin Kim, Jongjae Park, Hyeonwoo Kang, Seungryong Kim^†

arXiv preprint, 2026

A framework that improves virtual try-on by explicitly aligning person–garment correspondences through supervised query–key matching in Diffusion Transformers, achieving state-of-the-art garment-detail preservation.

arXiv · Code · Project Page

Experiences

Research Experience

Jan 2026 – Present
Undergraduate Researcher, CVLAB, KAIST AI · advised by Prof. Seungryong Kim
- Analyzing human–garment correspondence in Diffusion Transformer-based virtual try-on, and proposing the CORAL method via attention alignment.
- Researching a multi-view generation method that improves geometric consistency using a tracking module and attention alignment.
May 2025 – Jan 2026
CTO, TERAMIME (Startup)
- Led development of a lip-based speaker identification model using a GCN and facial landmarks.
- Built a generative-model-based liveness detection system; secured 8× NVIDIA H200 GPUs through the NIPA national GPU program.
Sep 2025 – Dec 2025
Project Leader, Advanced Machine Learning Project
- Developed a one-class liveness detection model using frequency-domain reconstruction and decomposition, leveraging the distribution of VAE reconstruction losses across low- and high-frequency bands.
Feb 2025 – Sep 2025
Undergraduate Researcher, Korea University AI-LAB
- Developed a pediatric emergency prediction model via multi-label classification on a BERT NLP model trained on St. Mary's Hospital data.
- Researched architectures addressing irregular time intervals in time-series EHR event sequences.

Honors & Awards

'20–'26 🎓 Top Student (Semester) ×6 & Honors Student ×1, Korea University
2025 🥇 1st Place, Seoul Investment Bridge IR Program — Seoul Regional SMEs & Startups Office Award
2025 🖥️ Selected, NIPA National GPU Program (8× NVIDIA H200)
2025 🌐 Finalist, WIPO Global Awards (World Intellectual Property Organization)
2025 🏆 CES Innovation Award, AI Category — TERAMIME
2024 🏅 Excellence Award, SK A-Dot Early Adopter Program
2023 🚀 Top 10 Team, NAVER AI RUSH 2023 — Generative AI Startup Track

Extracurricular Activities

Mar 2025 – Jul 2025
Team Member, Deep Learning Lecture Project
- Implemented a lip-reading model for four-digit number sequences using VTP and GRU, and wrote a research paper-style report.
Nov 2024 – Dec 2024
Team Member, SK A-Dot Early Adopter Activity
- Built a reading-promotion template on the Adot generative-AI platform using multi-agent systems and prompt engineering.
Aug 2024 – Dec 2024
Member, KAIROS Society
- Built a supplement-recommendation service with a self-constructed database and an LLM-based model, plus a RAG-based data-labeling and image-retrieval pipeline using LLMs and Pinecone.
Mar 2024 – Dec 2024
Member, EPOCH Society
- Built an ensemble model to predict movie success from film metadata, and analyzed youth culture via clustering on data from adolescents in multicultural families.
Sep 2023 – Dec 2023
Team Member, Naver Ambassador (NAVER AI RUSH 2023)
- Researched prompt-tuning techniques to implement an STT service based on HyperCLOVA X; selected as a Top 10 team in the Generative AI Startup Track.

Education

2020 – 2026
B.S., Korea University · Seoul, South Korea
Health & Environmental Convergence Science / AI Application · GPA 4.09 / 4.5
- Relevant coursework: Advanced Machine Learning, Machine Learning, Deep Learning, Natural Language Processing, Artificial Intelligence, Probability & Random Processes, Linear Algebra, Medical Image Processing, Data Structures.