Home

About Zhepei

Hello, this is Zhepei's page. I am an Applied Scientist at Adobe Firefly Applied Science & Machine Learning (ASML). I am particularly interested in the applications of machine learning in audio domain, including sound synthesis, sound recognition, source separation, speech enhancement, and multi-modal representation learning.

Before joining Adobe, I was an applied scientist at Amazon Web Services (AWS) for both full-time and several internship experiences, solving customer problems on real-time speech enhancement and enterprise search. I earned my Ph.D. degree from the Department of Computer Science at the University of Illinois Urbana-Champaign (UIUC) in 2023, advised by Professor Paris Smaragdis. Prior to my Ph.D. program, I obtained my B.S. in computer science from Harvey Mudd College in May 2018.

Feel free to checkout my academic CV , Google Scholar, and LinkedIn profile.

I'm also an enthusiastic fan of classical music, an amateur piano player, and a licensed bartender.

Selected Publications

  1. Unsupervised Improvement of Audio-Text Cross-Modal Representations Zhepei Wang, Yusuf Cem Subakan, Krishna Subramani, Junkai Wu, Tiago Tavares, Fabio Ayres, and Paris Smaragdis In IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA) 2023 [PDF] [Video] [Poster] [Code]
  2. A Framework for Unified Real-time Personalized and Non-Personalized Speech Enhancement Zhepei Wang, Ritwik Giri, Devansh Shah, Jean-Marc Valin, Michael M. Goodwin, and Paris Smaragdis In IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2023 [PDF] [Video] [Poster]
  3. Learning Representations for New Sound Classes With Continual Self-Supervised Learning Zhepei Wang, Yusuf Cem Subakan, Xilin Jiang, Junkai Wu, Efthymios Tzinis, Mirco Ravanelli, and Paris Smaragdis In IEEE Signal Processing Letters 2022 [PDF] [Poster] [Code]
  4. Semi-Supervised Singing Voice Separation With Noisy Self-Training Zhepei Wang, Ritwik Giri, Umut Isik, Jean-Marc Valin, and Arvindh Krishnaswamy In IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2021 [PDF] [Video] [Poster]
  5. Continual Learning of New Sound Classes Using Generative Replay Zhepei Wang, Yusuf Cem Subakan, Efthymios Tzinis, Paris Smaragdis, and Laurent Charlin In IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA) 2019 [PDF]

News

  • May 13, 2024
    I have joined Adobe Firefly ASML as an applied scientist.
  • Sep 25, 2023
    I have started my role as an applied scientist at AWS ChimeSDK Audio Science team.
  • Aug 17, 2023
    I have succesfully defended my thesis, “Data-Efficient Approaches for Audio Classification and Separation”. Slides are also available.
  • Jul 12, 2023