Wei-Ning Hsu

32-G442

77 Massachusetts Ave.

Cambridge, MA, 02139

Hi, I am a fifth year CS PhD student at Massachusetts Institute of Technology, working with my advisor James Glass at the Spoken Language Systems Group under CSAIL.

My research focuses on interpretable speech representation learning, using unlabeled, weakly-labeled, partially-labeled, and multi-modal data. It draws close connections with human speech perception, and provides a unified framework to leverage vast quantities of data in the wild. More importantly, it can facilitate numerous common speech tasks, such as robust speech recognition, expressive multi-speaker speech synthesis, speaker verification, speech enhancement, etc.

Over the past few years, I was also very fortunate to have the opportunities to collaborate many great researchers at many different places in the industry:

In Summer 2019, I was at Facebook AI Research (FAIR), New York, working with Awni Hannun, Gabriel Synnaeve, and Ann Lee on self-supervised speech recognition.
In Summer 2018, I was at Google Brain, Mountain View, working with Yu Zhang, Ron Weiss, Yuxuan Wang, Heiga Zen, and Yonghui Wu on multi-factor speech synthesis.
In Summer 2016, I was at Mitsubishi Electric Research Labs (MERL), Cambridge, working on speech separation and multi-channel speech recognition with John Hershey, Jonathan Le Roux, and Shinji Watanabe.

Before PhD, I received B.S. in Electrical Engineering from National Taiwan University in 2014. As an undergrad researcher, I worked with Prof. Lin-Shan Lee in Speech Processing Lab on spoken term detection, and with Prof. Hsuan-Tien Lin in Computational Learning Lab on active learning.