Xuan Shi (史璇)

I am a Ph.D. student in Electrical Engineering from University of Southern California, supervised by Prof. Shrikanth (Shri) S. Narayanan. I primarily focus my research on speech and audio processing, with specific interests in: speech production and articulatory knowledge analysis, music understanding.

I also have long-term collaborations with Dr. Erica Cooper, Dr. Xin Wang, and Prof. Junichi Yamagishi at National Institute of Informatics, Japan on musical instruments timbre analysis and music audio synthesis. Outside of USC and NII, I have spent several interning at research labs in academia and industry: Tencent, NLPR CASIA, and ByteDance.

Email  /  Google Scholar  /  GitHub  /  LinkedIn

profile photo
Publications & Preprints
A review of speech-centric trustworthy machine learning: Privacy, safety, and fairness
Tiantian Feng, Rajat Hebbar*, Nicholas Mehlman*, Xuan Shi*, Aditya Kommineni*, Shrikanth S. Narayanan
APSIPA Transactions on Signal and Information Processing, 2023.
(* equal contribution)
CAN KNOWLEDGE OF END-TO-END TEXT-TO-SPEECH MODELS IMPROVE NEURAL MIDI-TO-AUDIO SYNTHESIS SYSTEMS?
Xuan Shi, Erica Cooper, Xin Wang, Junichi Yamagishi, Shrikanth S. Narayanan
IEEE International Conference on Acoustic, Speech, and Signal Processing (ICASSP), 2023.
Creating musical features using multi-faceted, multi-task encoders based on transformers
Timothy Greer, Xuan Shi, Benjamin Ma, Shrikanth S. Narayanan
Scientific reports, Nature Publishing Group, 2023.
Use of Speaker Recognition Approaches for Learning and Evaluating Embedding Representations of Musical Instrument Sounds
Xuan Shi, Erica Cooper, Junichi Yamagishi
IEEE/ACM Transactions on Audio, Speech, and Language Processing, 2022. (Impact Factor: 3.919)
End-to-End Model for Speech Enhancement by Consistent Spectrogram Masking
Xingjian Du, Mengyao Zhu, Xuan Shi, Xinpeng Zhang, Wen Zhang, Jingdong Chen
CoRR, abs/1901.00295
End-to-End Residual CNN with L-GM Loss Speaker Verification System
Xuan Shi, Xingjian Du, Mengyao Zhu
IEEE International Conference on Digital Signal Processing (DSP), 2018.
Workshops
Detecting Poisoning Attacks against Speech Datasets using Variational Autoencoders
Nicholas Mehlman, Xuan Shi, Aditya Kommineni, Shrikanth S. Narayanan
3rd Symposium on Security and Privacy in Speech Communication, INTERSPEECH 2023
Unlocking Foundation Models for Privacy-Enhancing Speech Understanding: An Early Study on Low Resource Speech Training Leveraging Label-guided Synthetic Speech Content
Tiantian Feng, Digbalay Bose, Xuan Shi, Shrikanth S. Narayanan
3rd Symposium on Security and Privacy in Speech Communication, INTERSPEECH 2023
Speaker Recognition Adapted for Musical Instruments
Xuan Shi, Erica Cooper, Junichi Yamagishi
The Young Female Researchers in Speech Workshop (YFRSW), INTERSPEECH 2021
RepGN:Object Detection with Relational Proposal Graph Network
Xingjian Du*, Xuan Shi*, Risheng Huang
NeurIPS workshop: New In ML, 2019
(* equal contribution)

Patent
A Realtime Speech Enhancement Method
Xingjian Du, Mengyao Zhu, Xuan Shi
CN201810908839.3

Experience
  • Research Intern, NII Yamagishi Lab, Spring 2021 - Present
  • Research Intern, ByteDance AI-Lab, Summer 2020 - Winter 2020
  • Undergraduate Research Assistant, Institute of Automation, Chineses Acadamy of Sciences, Winter 2018 - Fall 2019
  • Research Intern, Tencent YouTu Lab, Fall 2018 - Summer 2019
  • Research Assistant, Shanghai University, Spring 2018 - Fall 2018

Last update: Oct. 2023      Template