I’m an Integrated student at Optimization and Statistical Inference Lab in Korea Advanced Institute of Science and Technology (KAIST), under the supervision of Prof. SEYOUNG YUN. I’m self-motivated student and love to solve many fun problems related to speech in real world setting.

My research interest includes:

Speech, TTS, Audio-Visual Generation, Audio-Visual Learning🎧, AVSR, CV

🔥 News

• 2025.03:  👀 Two papers have been archived on arXiv
• 2025.03:  🎉🎉 One paper is accept at TMLR
• 2025.01:  🎉🎉 One paper is accept at ICLR’25

📝 Publications

arXiv

• MAVFlow: Preserving Paralinguistic Elements with Conditional Flow Matching for Zero-Shot AV2AV Multilingual Translation
  • Sungwoo Cho, Jeongsoo Choi, Sungnyun Kim, Seyoung Yun

arXiv

• MoHAVE: Mixture of Hierarchical Audio-Visual Experts for Robust Speech Recognition
  • Sungnyun Kim, Kangwook Jang, Sangmin Bae, Sungwoo Cho, Seyoung Yun

ICLR 2025

• Multi-Task Corrupted Prediction for Learning Robust Audio-Visual Speech Representation
  • Sungnyun Kim, Sungwoo Cho, Sangmin Bae, Kangwook Jang, Seyoung Yun

TMLR

• FedDr+: Stabilizing Dot-regression with Global Feature Distillation for Federated Learning
  • Seongyoon Kim, Minchan Jeong, Sungnyun Kim, Sungwoo Cho, Sumyeong Ahn, Seyoung Yun

WACV 2023

• Phantom of Benchmark Dataset: Resolving Label Ambiguity Problem on Image Recognition in the Wild
  • Hyunhee Chung*, Kyung Ho Park*, Taewon Seo, Sungwoo Cho (*: equal contribution)

📝 Projects

• 2024.03 - 2024.12, ETRI, On-device Federated Learning , OSILAB
  • Development of adaptive lightweight edge-interconnected analysis technology with active instant response and rapid learning capability
• 2022.08 - 2023.03, Hankook Tire, Integrated prediction model using pattern extraction and pattern of tire image and code data, MLILAB
  • Complex modeling of tabular data and pattern embedding
• 2022.08 - 2022.12, EE488, DCASE 2020 Challenge(grade : A+)
  • Anomaly detection challenge of sound data
• 2022.04 - 2022.06, Weakly-Supervised Object Detector(Vehicle damage detection project), SOCAR AI team
  • Weakly supervised application to vehicle rental photo data
• 2022.02 - 2022.04, AI-based car wash request efficiency Check the robustness of the large classification model, SOCAR AI team
  • Development of deep learning-based car wash certification automation model(url)

📖 Educations

• 2024.03 - 2026.02 (now), Integrated Student (MS/PhD) in OSILAB. GSAI. KAIST.
• 2019.02 - 2024.02, B.E. in Electronic and Electrical Engineering. KAIST.

👨🏻‍🏫 Teaching Assistant

• KAIST AI707 Advanced Topics in Deep Reinforcement Learning, 2024 Fall

💻 Research Experiences

• 2022.08 - 2023.09, Research Intern, MLILAB, KAIST, South Korea.
• 2022.02 - 2022.08, Applied Scientist Intern, SOCAR, South Korea.
• 2021.12 - 2022.01, Research Student, ALIN-LAB, KAIST, South Korea.

💬 Other Works

• 2024.12, KAIST-KIST-Inria Workshop Student Organizer
  • French-Korean Workshop on AI, Networks and Communication, Programming Languages and Applied Mathematics
• 2021.07 - 2022.03, Founding Member at military community application, GUNINNI
  • Anonymous community app for soldiers
  • UI design manager
  • Responsible for business plan for start-up support
• 2021.03 - 2021.12, Industrial Design Department Student Council, KAIST
  • Department briefing session support
  • Conducting career interviews with seniors
  • Planning and proceeding with various TF projects