Scholar
Wei Kang
Google Scholar ID: dyIpw7kAAAAJ
Senior engineer, Xiaomi Corp.
Speech Recognition
Text to Speech
Follow
Homepage
↗
Google Scholar
↗
Citations & Impact
All-time
Citations
429
H-index
8
i10-index
7
Publications
18
Co-authors
9
list available
Contact
Email
wkang@pku.edu.cn
GitHub
Open ↗
Publications
1 items
OmniVoice: Towards Omnilingual Zero-Shot Text-to-Speech with Diffusion Language Models
2026
Cited
0
Resume (English only)
Academic Achievements
2025.02: CR-CTC accepted by ICLR 2025
2024.01: Zipformer accepted by ICLR 2024
2023.12: Libriheavy and PromptAsr accepted by ICASSP 2024
2023.05: Released a new version (stable version) of Zipformer
2023.05: Blank skipping for transducer and delay-penalized CTC accepted by Interspeech 2023
2023.02: Fast decoding for transducer, MVQ training, and Delay-penalized transducer accepted by ICASSP 2023
2022.11: Released the first version of Zipformer
2022.06: Started the sherpa, sherpa-onnx, sherpa-ncnn projects
2022.06: Pruned rnnt paper accepted by InterSpeech 2022
2022.05: Finished a reworked version of Conformer
2022.03: Finished pruned rnnt loss
2021.12: Decided to change direction from CTC/MMI to RNN-T models
2021.09: Released the first version of Icefall at InterSpeech 2021
Background
Currently a senior engineer at Xiaomi AI Lab, graduated from Peking University. Focusing on Speech, AI, and Machine Learning.
Miscellany
Personal interests and hobbies not mentioned
Co-authors
9 total
Zengwei Yao (姚增伟)
Machine Learning Engineer, Next-Gen Kaldi Team, Xiaomi Corp.
Fangjun Kuang
Xiaomi
Daniel Povey
Chief Speech Scientist, Xiaomi Corp.
Xiaoyu Yang
University of Cambridge
Yifan Yang
Shanghai Jiao Tong University, Tencent, Microsoft, Xiaomi
Mingshuang Luo
ICT, UCAS, Peng Cheng Lab
Piotr Żelasko
Principal Research Scientist @ Nvidia
Quandong Wang
Senior Speech Engineer, Xiaomi Corporation, Beijing, China
×
Welcome back
Sign in to Agora
Welcome back! Please sign in to continue.
Email address
Password
Forgot password?
Continue
Do not have an account?
Sign up