Deep Neural Networks for Acoustic Modeling in Speech Recognition

Geoffrey Hinton, Li Deng, Dong Yu, Abdel-rahman Mohamed, Navdeep Jaitly, Andrew Senior, Vincent Vanhoucke, Patrick Nguyen, Tara Sainath George Dahl, and Brian Kingsbury

Abstract

Most current speech recognition systems use hidden Markov models (HMMs) to deal with the temporal variability of speech and Gaussian mixture models (GMMs) to determine how well each state of each HMM fits a frame or a short window of frames of coefficients that represents the acoustic input. An alternative way to evaluate the fit is to use a feed-forward neural network that takes several frames of coefficients as input and produces posterior probabilities over HMM states as output. Deep neural networks (DNNs) that have many hidden layers and are trained using new methods have been shown to outperform GMMs on a variety of speech recognition benchmarks, sometimes by a large margin. This article provides an overview of this progress and represents the shared views of four research groups that have had recent successes in using DNNs for acoustic modeling in speech recognition.

Details

Publication typeArticle
Published inIEEE Signal Processing Magazine
URLhttp://psych.stanford.edu/~jlm/pdfs/Hinton12IEEE_SignalProcessingMagazine.pdf
Pages82-97
Volume29
Number6
> Publications > Deep Neural Networks for Acoustic Modeling in Speech Recognition