Unsupervised Speech Recognition
Alexei Baevski, Wei-Ning Hsu, Alexis Conneau, Michael Auli
People
Alexei Baevski
Research Engineer
I work at Facebook AI Research on NLP problems and toolkits such as as Fairseq. Previously, I worked on Facebook Search, specifically on spell correction and query suggestion problems. Before that, I was at a financial risk management start-up where I worked as an engineer and a manager. I hold a Bachelor’s degree in computer science from University of Toronto.
Interests
NLP on its own and combined with other fields like vision (e.g. visual question answering), conversational systems (chatbots, conversational search, assistants), recommender systems and performance optimizations for existing algorithms
Latest Publications
NeurIPS - 2021
wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations
Alexei Baevski, Henry Zhou, Abdelrahman Mohamed, Michael Auli
Effectiveness of self-supervised pre-training for ASR
Alexei Baevski, Abdelrahman Mohamed
vq-wav2vec: Self-Supervised Learning of Discrete Speech Representations
Alexei Baevski, Steffen Schneider, Michael Auli
Cloze-driven Pretraining of Self-attention Networks
Alexei Baevski, Sergey Edunov, Yinhan Liu, Luke Zettlemoyer, Michael Auli
wav2vec: Unsupervised Pre-training for Speech Recognition
Steffen Schneider, Alexei Baevski, Ronan Collobert, Michael Auli