Curriculum Vitae
The short version. Download the full PDF ↗
Speech and audio understanding models for Gemini: core model capabilities and safety alignment for audio-in and audio-out launches. Contributed to Gemini 3.5 Transcribe, Speak to Window (Gemini for macOS), Rambler in Gboard (Pixel 11), and Gemini 2.5 TTS on Vertex AI.
Manager: Dr. Quan Wang
Founding team member. Contributed to Trillion-7B and Tri-21B, multilingual LLMs for practical, real-world applications. Worked on model architecture and training infrastructure.
Robust audiovisual representation learning under missing modalities, recovering absent information when only audio or video is available.
Manager: Dr. Anurag Kumar
Advisor: Prof. Yong Man Ro
Thesis: Delving into Human Speech Understanding through Multimodal Representation Learning
- Organizer, Audio-Visual Generation and Learning Workshop, ECCV 2024 & 2026
- Onsite Coordinator, SANE 2025
- Organizing Committee, KAIST AI World Cup 2018
- PRHLT, Universitat Politècnica de València (2025)
conferences
- AAAI2025–2027
- AAAI AI Alignment Track2026
- BMVC2026
- CVPR2024, 2025
- ECCV2024
- ICASSP2024–2027
- ICLR2024–2027
- ICCV2023, 2025
- ICML2021–2025
- NeurIPS2020, 2022–2024, 2026
- NeurIPS Datasets & Benchmarks2025
- NeurIPS Workshop2024
- WACV2026, 2027
journals
- IEEE TIP2025
- IEEE TASLP2024
- IEEE TCSVT2023, 2024, 2026
- IEEE TMM2023
- EE474 Introduction to Multimedia (2020–2023)
- EE837 Multimedia Processing and Learning (2022)
- EE534 Pattern Recognition (2021)