Jiawei Du is from mainland China and has been studying in Taiwan for several years. Chinese speakers can call me by my Chinese name, 杜嘉煒 or 杜嘉炜, while English speakers can simply call me Jiawei or Gavin.

Jiawei is currently a third-year master student in the Department of Computer Science and Information Engineering at National Taiwan University. He is an active researcher at the Multimedia Information Retrieval Lab (MIRLab), supervised by Prof. Jyh-Shing Roger Jang in the areas of speech processing and machine learning. He obtained his B.S. degree from the Department of Information and Telecommunications Engineering at Ming Chuan University, under the mentorship of Prof. Shu-Yin Chiang in Robotics.

Jiawei’s research focuses on speech technology and its applications, including audio-centric generation models, neural audio codecs, and audio deepfake detection and localization. These efforts aim to help distinguish truth from falsehood in the era of AIGC proliferation.

Jiawei is planning to pursue a Ph.D. with a primary focus on research in speech processing and machine learning. To better prepare for this path, he is actively seeking international internship opportunities. If you are aware of any suitable Ph.D. programs or relevant internships, please don’t hesitate to get in touch with him.

News

[05/2025] Completed a 4-month research internship at Samsung Research China (Beijing), focusing on neural audio codec.

[02/2025] One paper accepted to Interspeech 2025.

[01/2025] Submitted preprint: CodecFake+: A Large-Scale Codec-based Deepfake Speech Dataset.

[09/2024] One paper accepted to APSIPA ASC 2024.

[08/2024] Three papers accepted to SLT 2024.

[06/2024] One paper accepted to Interspeech 2024.

[06/2024] Ranked 13/49 teams in the Singing Voice Deepfake Detection Challenge.

[02/2024] Submitted preprint: EMO-SUPERB: An In-depth Look at Speech Emotion Recognition.

[05/2023] Ranked 3rd in the DCASE Challenge Task 6b.