CV
Basics
| Name | Sangmin Lee |
| Label | Deep Learning Researcher |
| sangmin_lee@dsp.yonsei.ac.kr | |
| Url | https://sanghyang00.github.io |
| Summary | PhD candidate at Yonsei University specializing in multilingual speech signal processing and speech large language models. Experienced in massively multilngual speech and NLP with a strong publication record. |
Education
Work
-
2024.03 - 2025.02 Teaching Assistant
Yonsei University, DSP&AI Lab
Served as a Teaching Assistant for various undergraduate courses in the Department of Electrical and Electronic Engineering.
- EEE2060: Signals and Systems (Fall 2024)
- EEE4610: Electrical and Electronic Engineering Capstone Design (Fall 2024)
- EEE4420: Digital Signal Processing (Spring 2024)
-
2023.08 - 2023.12 Undergraduate Research Intern
Korea University, CV Lab
Undergraduate Research Assistant in Speech-Driven 3D Facial Generation under the supervision of Prof. Seungryong Kim.
-
2021.10 - 2022.02 Full Stack Developer
Neoforce Incorporation
Implemented 3D Medical Imaging Applications with Python.
International Publications
-
2025.09.01 UniCoM: A Universal Code-Switching Speech Generator
Findings of EMNLP 2025 (Poster)
Authors: *Sangmin Lee*, Woojin Chung, Seyun Um, and Hong-Goo Kang
-
2025.01.01 LAMA-UT: Language Agnostic Multilingual ASR Through Orthography Unification and Language-Specific Transliteration
AAAI 2025 (Oral)
Authors: *Sangmin Lee*, Woojin Chung, and Hong-Goo Kang
Preprints
-
2025.12.01 AEGIS: Awareness-Enhanced Guidance for Iterative Safeguard
To be appear
Authors: Kyungwon Park, *Sangmin Lee*, Heejae Chon, and Hyungu Kang
-
2025.10.01 SAGE-LD: Towards Scalable and Generalizable End-to-End Language Diarization via Simulated Data Augmentation
arXiv 2025
Authors: *Sangmin Lee*, Woongjib Choi, Jihyun Kim, and Hong-Goo Kang
-
2025.10.01 UniverSR: Unified and Versatile Audio Super-Resolution via Vocoder-Free Flow Matching
arXiv 2025
Authors: Woongjib Choi, *Sangmin Lee*, Hyeongseob Lim, and Hong-Goo Kang
-
2024.10.01 Talk3d: High-fidelity talking portrait synthesis via personalized 3d generative prior
arXiv 2024
Authors: Jaehoon Ko, Kyusun Cho, Joungbin Lee, Heeji Yoon, *Sangmin Lee*, Sangjun Ahn, and Seungryong Kim
Projects
-
2023.03 - 2023.06
Music Genre Classifier & Recommendation System Project
Developed a model that classifies music genres and recommends songs based on metadata information.
- Conducted as a team project under Korea University AI Club (AIKU)
-
2023.05 - 2023.06
Music Rating Archive Web Project
Built a web application to organize and view music metadata along with personal ratings.
- Completed as an assignment for the undergraduate Database course (COSE37102)
-
2023.06 - 2023.07
Time Series Sensor Data Anomaly Detection using LSTM AutoEncoder
Developed a model to predict errors in factory processes using sensor data.
- Performed as a commissioned project for an acquaintance
-
2023.07 - 2023.08
AI Lyricist Project
Created a model that generates appropriate lyrics from music waveforms.
- Conducted as a team project under Korea University AI Club (AIKU)
-
2023.09 - 2023.11
2023 Korean AI Contest Track 1 (Speech Recognition)
Developed a Korean speech recognition system specialized for regional dialects, elderly, and children.
- Ranked 7th out of 302 teams (Team AIKUDIO)
Awards
-
2023.11.13
National Information Society Agency Prize
President of NIA
Awarded for outstanding performance as the 7th place team in the 2023 Korean AI Contest Track 1.
Extracurricular Activities
-
2024.01 - present Seoul, South Korea
-
2023.01 - 2023.12 Seoul, South Korea
-
2020.04 - 2021.10 Incheon, South Korea
Sergeant, Repair Parts and Tool Supply Specialist
Republic of Korea Army
Completed mandatory military service.