Welcome to the InfoSci Platform
Reference Hub1
Applications of Kernel Theory to Speech Recognition

Applications of Kernel Theory to Speech Recognition

Joseph Picone, Aravind Ganapathiraju, Jon Hamaker
Copyright: © 2007 |Pages: 22
ISBN13: 9781599040424|ISBN10: 1599040425|ISBN13 Softcover: 9781599040431|EISBN13: 9781599040448
DOI: 10.4018/978-1-59904-042-4.ch009
Cite Chapter Cite Chapter

MLA

Picone, Joseph, et al. "Applications of Kernel Theory to Speech Recognition." Kernel Methods in Bioengineering, Signal and Image Processing, edited by Gustavo Camps-Valls, et al., IGI Global, 2007, pp. 224-245. https://doi.org/10.4018/978-1-59904-042-4.ch009

APA

Picone, J., Ganapathiraju, A., & Hamaker, J. (2007). Applications of Kernel Theory to Speech Recognition. In G. Camps-Valls, J. Rojo-Alvarez, & M. Martinez-Ramon (Eds.), Kernel Methods in Bioengineering, Signal and Image Processing (pp. 224-245). IGI Global. https://doi.org/10.4018/978-1-59904-042-4.ch009

Chicago

Picone, Joseph, Aravind Ganapathiraju, and Jon Hamaker. "Applications of Kernel Theory to Speech Recognition." In Kernel Methods in Bioengineering, Signal and Image Processing, edited by Gustavo Camps-Valls, Jose Luis Rojo-Alvarez, and Manel Martinez-Ramon, 224-245. Hershey, PA: IGI Global, 2007. https://doi.org/10.4018/978-1-59904-042-4.ch009

Export Reference

Mendeley
Favorite

Abstract

Automated speech recognition is traditionally defined as the process of converting an audio signal into a sequence of words. Over the past 30 years, simplistic techniques based on the design of smart feature-extraction algorithms and physiological models have given way to powerful statistical methods based on generative models. Such approaches suffer from three basic problems: discrimination, generalization, and sparsity. In the last decade, the field of machine learning has grown tremendously, generating many promising new approaches to this problem based on principles of discrimination. These techniques, though powerful when given vast amounts of training data, often suffer from poor generalization. In this chapter, we present a unified framework in which both generative and discriminative models are motivated from an information theoretic perspective. We introduce the modern statistical approach to speech recognition and discuss how kernel-based methods are used to model knowledge at each level of the problem. Specific methods discussed include kernel PCA for feature extraction and support vector machines for discriminative modeling. We conclude with some emerging research on the use of kernels in language modeling.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.