Highlights:
• Architect of Voci’s automatic speech recognition(ASR) engine (serve >>50+M hrs/yr),
• Manager of an ASR R&D group which develop acoustic/language model for Voci’s ASR engine,
• Early employee of Voci (Employee #9, acquired in 2020 by Medallia),
• Admin of Facebook’s Largest AI/DL Group (525k+ members),
• Past: Maintainer of Sphinx. Early employee of Scanscout (Employee #4, acquired in 2010 by TremorVideo).
My personal mission statement: apply human language technology and machine learning to improve everyone’s life.
My past research life:
• Research staff at BBN, Speechworks and Scanscout.
• Coauthor of one of the best papers in a prestigious international conference.
• Volunteer researcher at MGH, with one publication.
Skills:
Programming: C and Python.
Speech Recognition-Related: Architecture of Speech Recognition Systems (Decoder+Trainer), Very Fast Speech Recognition, Keyword Spotting, Speech-based Topic/Language/Emotion/Gender Classification, Robust Speech Recognition.
General Machine Learning-Related: Application Of Machine Learning Algorithms (Regression, SVM, GMM), Sentiment classification, Information retrieval.
Deep Learning-Related (i.e. DNN, CNN, RNN and variants):
Speech recognition: Deep learning in acoustic and language modeling.
Administration: knowledgeable in setup and install multiple deep learning tools.
Toolkit Expertise (ASR): Sphinx (2,3,4 and pocketsphinx), Kaldi, HTK, Julius, Speechworks (