معرفی
Xiaofei LI is an Assistant Professor at Westlake University, Hangzhou, China, a position he has held since March 2020. Prior to this, he was a post-doctoral researcher and later a starting research scientist at INRIA Grenoble Rhône-Alpes, France, from February 2014 to at least 2019. He is affiliated with the PERCEPTION team and has contributed to major projects such as the ERC VHIA and EARS projects, as well as industry collaboration with Samsung Electronics.
- Education
- Bachelor of Electronic Information, Beijing Institute of Machinery, 2007
- PhD in Electronics, Peking University, 2007–2013
His research focuses on speech and audio signal processing, particularly in challenging real-world environments. Key areas include multi-microphone speech processing for sound source localization, separation, and dereverberation; single-microphone signal processing for noise estimation, voice activity detection, and speech enhancement; and audio-visual fusion for speaker diarization and tracking. He employs advanced techniques such as LSTM networks, Bayesian inference, convolutive transfer functions, and deep learning in the STFT domain.
His recent publications (2017–2019) demonstrate a strong trend toward multichannel speech enhancement, online speaker localization and tracking, and audio-visual integration. His work frequently appears in top-tier IEEE journals such as IEEE/ACM Transactions on Audio, Speech and Language Processing and IEEE Transactions on Pattern Analysis and Machine Intelligence. He has developed and evaluated algorithms using real datasets like LOCATA and AVDIAR, emphasizing robustness in reverberant and multi-speaker environments.
Scientific Awards and Recognition
- No specific awards are mentioned in the provided text.
Advising and Grants
- Currently serves as an Assistant Professor, implying student supervision and research advising, though no specific students are named.
- Research has been funded by prestigious grants including the European Research Council (ERC) Advanced Grant VHIA #340113 and the EU-FP7 STREP project EARS (#609465).
Labs and Research Teams
- PERCEPTION team, INRIA Grenoble Rhône-Alpes, France
- Involved in the EARS, ERC VHIA, and Samsung Electronics Lito projects
- Contributor to the development of the AVDIAR and AVTrack-1 datasets
