● Ph.D. major in Information and Computer Science.● Ph.D. minor in Global Resource Management---Japan Leading Program on Humanistic and Social Science.● Research Interests on speech recognition, machine learning, deep learning, language technologies for education, computational linguistics, and spoken language systems.● 8+ years experience on speech recognition, machine learning and language technologies for education ● Experienced in multi-modal systems, signal processing, data analysis and system development.● Experience with the programming language: Python, Perl, Matlab, UNIX shell scripting, LATEX.● Experience with Research Tools: TensorFlow, Kaldi, Theano, CMUSphinx, HTK, Julius, CMU SLM, SRILM, T3, OpenFST, SprinTra.
-
Speech And Machine Learning ScientistBlue Canoe Feb 2018 - Feb 2021Bellevue, Washington, UsDesign a game-based framework for data collection and labeling to improve automatic speech recognition (ASR) performance. Analyze and validate phoneme-level insertion, deletion and substitution issues in accented English across more than 86 languages and over 150 countries. Train machine learning models to offer multiple levels of pronunciation feedback.Develop customized classifiers to identify the unique pronunciation challenges of learners with different first languages. -
Research AssistantDoshisha University Apr 2016 - Mar 2017Kyoto, Kyoto, JpLead a team of graduate and undergraduate students to develop a conversation based, computer-assisted language learning system to support Japanese students in learning other languages. Design the system with conversation scenarios tailored for language learners. Collaborate on data collection. Built a tolerant automatic speech recognition ASR system to recognize Japanese accented English. -
Research FellowNanyang Technological University 2017 - 2017Singapore, Singapore, SgBuild a Convolutional Neural Network(CNN) based speaker diarization system to classify speech from teachers, children and overlapping speech for classroom analysis. Perform speech boundary detection using the Generalized Likelihood Ratio-Bayesian Information Criterion (GLR-BIC) to increase accuracy from 68% to 85%. -
Visiting Graduate ResearcherUniversity Of California, Los Angeles May 2016 - Sep 2016Los Angeles, Ca, UsInvestigate the phenomena of acoustic variations in vowels and consonants. Use a data-driven method to automatically generate a more effective phonetic units (phoneme set) for training an ASR system capable of recognizing K-12 children's speech. Combine Transfer Learning (TL) with Multiple Task Learning (MTL) to develop a tolerant deep neural network based acoustic model. -
Collaborative ResearcherNict May 2015 - Mar 2016Koganei, Tokyo, JpInvestigate F0 (fundamental frequency) related tonal features in speech recognition for tonal language. Analyze F0 contours using signal processing methods to reconstruct F0, aiming to remove irrelevant components of tonal information for the speech recognition of tonal language. Use reconstructed features, which include prominent lexical information, to train a DNN-based acoustic model for Mandarin speech recognition. -
Visiting Graduate ResearcherUniversity Of Helsinki Jun 2014 - Jun 2014Helsinki, FiWork on multi-modal and human-robot interactions for conversing using Wikipedia content in Finnish, English, and Japanese. Design the robot's activity functions, including an ASR module, a gaze recognition module, and a gesture recognition module. Build a robot capable of switching between English and Japanese language for talking, utilizing the Wikipedia database.
Xiaoyun Wang Education Details
-
Doshisha UniversityInformation And Computer Science -
Doshisha UniversityInformation And Computer Science
-
Yamanashi Eiwa CollegeHumanities/Humanistic Studies (Minor: Information Science)
Frequently Asked Questions about Xiaoyun Wang
What is Xiaoyun Wang's role at the current company?
Xiaoyun Wang's current role is Speech and Machine Learning Scientist.
What schools did Xiaoyun Wang attend?
Xiaoyun Wang attended Doshisha University, Doshisha University, Yamanashi Eiwa College.
Free Chrome Extension
Find emails, phones & company data instantly
Aero Online
Your AI prospecting assistant
Select data to include:
0 records × $0.02 per record
Download 750 million emails and 100 million phone numbers
Access emails and phone numbers of over 750 million business users. Instantly download verified profiles using 20+ filters, including location, job title, company, function, and industry.
Start your free trial