Scholar
Daniel Povey
Google Scholar ID: y_-5FWAAAAAJ
Chief Speech Scientist, Xiaomi Corp.
Citations & Impact
Citations
31,458
H-index
57
i10-index
119
Publications
20
Co-authors
52
list available
Contact
No contact links provided.
Publications
Resume (English only)
Academic Achievements
- Submitted a paper titled 'Zipformer: A faster and better encoder for automatic speech recognition' to ICLR 2024.
- Contributed to several research projects presented at ICASSP 2024, including 'Libriheavy: a 50,000 hours ASR corpus with punctuation casing and context', 'PromptASR for contextualized ASR with controllable style', and 'Towards Universal Speech Discrete Tokens: A Case Study for ASR and TTS'.
- Co-authored numerous papers in the field of Automatic Speech Recognition (ASR) spanning from 2023 to 2021, featured in key international conferences such as Interspeech and ICASSP.
- Research topics extensively cover various aspects of ASR technology, including weakly supervised learning, semi-supervised learning, low-latency streaming ASR solutions, etc.
- Also explored areas like keyword search system construction, knowledge distillation methods, and improved acoustic modeling strategies.
