AI Paper Talks: Self-Supervised Speech Representation Learning
Join Avneet Singh as he discusses the paper “Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units”
The paper “Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units” introduces how Hubert applies novel self-supervision techniques to learn good representations and sequencing relationships among human speech audio segments.
Hubert then trains a near billion parameter model to achieve similar or better results than earlier top automatic speech recognition models, like wav2vec 2.0.
Bio: Avneet Singh started working on AI 20 years ago at CMI – Prince University using AI techniques to combat catastrophic climate change. He later worked in manufacturing R&D, real estate prediction, and search engine and ads at Bing Microsoft. Currently at Nextiva, Avneet is working on Speech Recognition, Reinforcement Learning and NLP problems to contribute to Nextiva’s goal to help build a stronger connection between businesses and their customers.
Nextiva Artificial Intelligence Paper Talks are technical discussions based on recent research about Natural Language Processing (NLP) or Artificial Intelligence (AI). This event will include a 30-minute presentation and 15 minutes of open discussion.