Dong Yu

Career

Yu joined Microsoft in 1998 and became a principal researcher at Microsoft Research. In 2017 Yu joined Tencent America and worked at Tencent AI Lab as a distinguished scientist and vice general manager until 2026. Yu is now VP of AI Foundations at Capital One and chief scientific advisor of Transsion.¹

In professional service, Yu was a member of the IEEE Speech and Language Processing Technical Committee from 2013 to 2018, its vice-chair in 2020 and its chair in 2021–2022, and was a technical co-chair of ICASSP 2021. Yu has been an associate editor and, for 2025–2027, a senior area editor of the IEEE/ACM Transactions on Audio, Speech, and Language Processing.¹

Research

Yu's research covers deep learning for speech processing and recognition, large language models and multimodal agentic systems. With George Dahl, Li Deng and Alex Acero, Yu co-authored the 2012 paper proposing context-dependent pre-trained deep neural network and hidden Markov model hybrids for large vocabulary speech recognition, which received the IEEE Signal Processing Society Best Paper Award in 2013.¹²

Yu wrote two monographs with Li Deng: "Deep Learning - Methods and Applications" (2014) and "Automatic Speech Recognition - A Deep Learning Approach" (Springer, 2014). Recent invited talks, such as an Interspeech 2025 survey talk, focus on building conversational agents.⁴⁵¹

Work and contributions

• Computational Network Toolkit (CNTK) Published research

  First author of "An Introduction to Computational Networks and the Computational Network Toolkit" (MSR-TR-2014-112), which introduces computational networks and describes CNTK, a GPU and CPU toolkit for building models such as DNNs, CNNs, RNNs, and LSTMs.³

• Context-dependent pre-trained deep neural networks for speech recognition Published research

  Co-author, with George Dahl, Li Deng, and Alex Acero, of the 2012 IEEE Transactions on Audio, Speech, and Language Processing paper proposing a pre-trained DNN-HMM hybrid architecture for large vocabulary speech recognition; the Microsoft Research page notes the IEEE SPS 2013 Best Paper Award.²

• Deep Learning: Methods and Applications Published research

  Co-author with Li Deng of this 2014 book (MSR-TR-2014-21, published by Microsoft) surveying deep learning methods and their applications to speech, language, vision, and other tasks.⁴

• Automatic Speech Recognition: A Deep Learning Approach Published research

  Co-author with Li Deng of this 2014 Springer monograph, whose chapters cover deep neural network-hidden Markov model hybrid systems and methods to speed up their training and decoding.⁵

Timeline

1. 1998Joined Microsoft, later becoming a principal researcher at Microsoft Research.¹
2. Jan 2012Co-authored the context-dependent pre-trained deep neural network paper for large vocabulary speech recognition.²
3. May 2014Published "Deep Learning - Methods and Applications" with Li Deng.⁴
4. Nov 2014Published the Springer monograph "Automatic Speech Recognition - A Deep Learning Approach" with Li Deng.⁵
5. 2017Joined Tencent America and Tencent AI Lab as a distinguished scientist.¹
6. 2021Became chair of the IEEE Speech and Language Processing Technical Committee and served as technical co-chair of ICASSP 2021.¹
7. Aug 21, 2025Gave an invited survey talk on conversational agents at Interspeech 2025.¹
8. 2026Concluded work at Tencent AI Lab as a distinguished scientist and vice general manager.¹

Other documented roles

Roles the sources document that are not confirmed as current.

• Distinguished Scientist and Vice General Manager (until 2026), Tencent AI Lab¹
• Principal Researcher (at Microsoft from 1998 until 2017), Microsoft Research¹

Honors and recognition

• ACM Fellow · Association for Computing Machinery¹
• IEEE Fellow · Institute of Electrical and Electronics Engineers¹
• ISCA Fellow · International Speech Communication Association¹
• IEEE SPS Claude Shannon-Harry Nyquist Technical Achievement Award · IEEE Signal Processing Society, 2025¹
• IEEE Signal Processing Society Best Paper Award · IEEE Signal Processing Society, 2013¹²
• NAACL Best Long Paper Award · North American Chapter of the Association for Computational Linguistics, 2021¹
• EMNLP Outstanding Paper Award · Conference on Empirical Methods in Natural Language Processing, 2023¹

Where they work

• Capital One 1 person in the wiki

Related people

• Jonathan CohenLanguage models · Speech and audio
• Ronan CollobertML frameworks and libraries · Speech and audio
• Gabriel SynnaeveLanguage models · Speech and audio
• Michael AuliSpeech and audio · Language models
• Ju-Chieh ChouSpeech and audio · Language models
• Shinji WatanabeSpeech and audio · ML frameworks and libraries
• Yu-An ChungSpeech and audio · Language models
• Lewis TunstallLanguage models · ML frameworks and libraries
• Alec RadfordLanguage models · Speech and audio
• Laurent SartranLanguage models · ML frameworks and libraries