Senior Applied Scientist at NVIDIA
E-mail: see my resume
Thesis
Google Scholar
YouTube Channel
| Biography |
-
I am currently working in NVIDIA's NeMo Team, supervised by Boris Ginsburg. Before joining NVIDIA, I worked in Google's Speech Team under Bhuvana Ramabhadran from September 2019 to October 2021, after getting my Ph.D. degree in Computer Science from the Johns Hopkins University, working in the Center for Language and Speech Processing (CLSP) under former JHU Prof. Daniel Povey and Prof. Sanjeev Khudanpur.
-
I received my B.S. in Software Engineering in 2012 from School of Software Engineering at Shanghai Jiaotong University in Shanghai, China. From 2012 to 2013, I worked with Professor Kai Yu in SJTU Speech Lab.
| Education |
- Ph.D. in Computer Science
(Sept 2013 - Sept 2019)
Department of Computer Science
Johns Hopkins University, MD, USA
- B.S. in Software Engineering
(Sept 2008 - Jun 2012)
School of Software Engineering
Shanghai Jiaotong University, Shanghai, China
| Work Experience |
- Senior Applied Scientist
NVIDIA (Nov 2021 - Present)
I do research in the NeMo team. My supervisor is Boris Ginsburg.
- Software Engineer
Google Inc, New York City, NY (Sept 2019 - Oct 2021)
I worked in the speech team, doing research to help Google's speech recognition models, supervised by Bhuvana Ramabhadran.
- Research Intern
Spoken Communications, Seattle, WA (May 2017 - Aug 2017)
I incorporated support for TensorFlow-based language model rescoring in Kaldi.
- Research Intern
Google Inc., New York City, NY (May 2015 - Aug 2015)
I worked with Cyril Allauzen and Michael Riley on improving contextual language modeling for Google speech recognition, using finite state methods implemented with OpenFst and OpenGrm.
- Research Assistant
Center for Language and Speech Processing (Sept 2013 - Sept 2019)
I worked with Dan Povey and Sanjeev Khudanpur on speech recognition, and contributed to the Kaldi project.
- Research Assistant
Speech Lab at Shanghai Jiaotong University (July 2012 - July 2013)
I worked with Kai Yu on speech recognition, speech synthesis and human-computer iteractions.
| Teaching Experience |
- Teaching Assistant
Johns Hopkins University (Spring Semester 2017)
Course: Information Extraction
Instructor: Sanjeev Khudanpur
- Teaching Assistant
Johns Hopkins University (Spring Semester 2016)
Course: Information Theory
Instructor: Sanjeev Khudanpur
- Teaching Assistant
Johns Hopkins University (Fall Semester 2015)
Course: Introduction to Machine Learning
Instructor: Ilya Shpitser
| Selected Publications |
- Shuoyang Ding, Hainan Xu, Philipp Koehn
Saliency-driven Word Alignment Interpretation for Neural Machine Translation
(WMT 2019) - Hainan Xu, Shuoyang Ding, Shinji Watanabe
Improving End-to-end Speech Recognition with Pronunciation-assisted Sub-word Modeling
(ICASSP 2019) - Hainan Xu, Tongfei Chen, Dongji Gao, Yiming Wang, Ke Li, Nagendra Goel, Yishay Carmiel, Daniel Povey, Sanjeev Khudanpur
A Pruned Rnnlm Lattice-rescoring Algorithm for Automatic Speech Recognition
(ICASSP 2018) - Hainan Xu, Ke Li, Yiming Wang, Jian Wang, Shiyin Kang, Xie Chen, Daniel Povey, Sanjeev Khudanpur
Neural Network Language Modeling with Letter-based Features and Importance Sampling
(ICASSP 2018) - Hainan Xu, Guoguo Chen, Daniel Povey and Sanjeev Khudanpur.
Modeling Phonetic Contexts with Non-random Forests for Speech Recognition
(Interspeech 2015)
Super easy-to-follow slides Conference Poster - Hainan Xu, Philipp Koehn.
Zipporah, an Automatic Data-cleaning System for Noisy Web-crawled Parallel Corpora
(EMNLP 2017) - Huda Khayrallah, Hainan Xu, Philipp Koehn.
The JHU Parallel Corpus Filtering Systems for WMT 2018
(WMT workshop of EMNLP 2018) - Yiming Wang, David Snyder, Hainan Xu, Vimal Manohar, Phani Sankar Nidadavolu, Daniel Povey, Sanjeev Khudanpur
The JHU ASR System for VOiCES from a Distance Challenge 2019
(Interspeech 2019) - Hainan Xu, Yuchen Fan, Kai Yu
Development of the 2012 SJTU HVR System
(14th ACM International Conference on Multimodal Interaction) - Guoguo Chen, Hainan Xu, Minhua Wu, Daniel Povey and Sanjeev Khudanpur.
Pronunciation and Silence Probability Modeling for ASR
(Interspeech 2015) - Kai Yu, Hainan Xu
Cluster Adaptive Training with Factorized Decision Trees for Speech Recognition
(Interspeech 2013) - Ke Li, Hainan Xu, Yiming Wang, Dan Povey and Sanjeev Khudanpur
Recurrent Neural Network Language Model Adaptation for Conversational Speech Recognition
(Interspeech 2018) - Zhehuai Chen, Justin Luitjens, Hainan Xu, Yiming Wang, Daniel Povey, Sanjeev Khudanpur
A GPU-based WFST Decoder with Exact Lattice Generation
(Interspeech 2018 ) - Yiming Wang, Vijayaditya Peddinti, Hainan Xu, Xiaohui Zhang, Daniel Povey, Sanjeev Khudanpur
Backstitch: Counteracting Finite-sample Bias via Negative Steps
(Interspeech 2017)
- Szu-Jui Chen, Aswin Shanmugam Subramanian, Hainan Xu, Shinji Watanabe
Building state-of-the-art distant speech recognition using the CHiME-4 challenge with a setup of speech enhancement baseline
(Interspeech 2018 ) - Daniel Povey, Gaofeng Cheng, Yiming Wang, Ke Li, Hainan Xu, Mahsa Yarmohamadi, Sanjeev Khudanpur
Semi-Orthogonal Low-Rank Matrix Factorization for Deep Neural Networks
(Interspeech 2018 ) - Jan Trmal, Matthew Wiesner, Vijayaditya Peddinti, Xiaohui Zhang, Pegah Ghahremani, Yiming Wang, Vimal Manohar, Hainan Xu, Daniel Povey, Sanjeev Khudanpur
The Kaldi OpenKWS System: Improving Low Resource Keyword Search
(Interspeech 2017)