Share Email Print
cover

Proceedings Paper

Continuous speech recognition based on convolutional neural network
Author(s): Qing-qing Zhang; Yong Liu; Jie-lin Pan; Yong-hong Yan
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Convolutional Neural Networks (CNNs), which showed success in achieving translation invariance for many image processing tasks, are investigated for continuous speech recognitions in the paper. Compared to Deep Neural Networks (DNNs), which have been proven to be successful in many speech recognition tasks nowadays, CNNs can reduce the NN model sizes significantly, and at the same time achieve even better recognition accuracies. Experiments on standard speech corpus TIMIT showed that CNNs outperformed DNNs in the term of the accuracy when CNNs had even smaller model size.

Paper Details

Date Published: 6 July 2015
PDF: 6 pages
Proc. SPIE 9631, Seventh International Conference on Digital Image Processing (ICDIP 2015), 963121 (6 July 2015); doi: 10.1117/12.2197152
Show Author Affiliations
Qing-qing Zhang, Key Lab. of Speech Acoustics and Content Understanding (China)
Yong Liu, Key Lab. of Speech Acoustics and Content Understanding (China)
Jie-lin Pan, Key Lab. of Speech Acoustics and Content Understanding (China)
Yong-hong Yan, Key Lab. of Speech Acoustics and Content Understanding (China)


Published in SPIE Proceedings Vol. 9631:
Seventh International Conference on Digital Image Processing (ICDIP 2015)
Charles M. Falco; Xudong Jiang, Editor(s)

© SPIE. Terms of Use
Back to Top