Adaptation in statistical pattern recognition using tangent vectors

Author

Keysers, Daniel ; Macherey, Wolfgang ; Ney, Hermann ; Dahmen, Jörg

Author_Institution

Dept. Comput. Sci., Aachen-Univ. of Technol., Aachen, Germany

Volume

26

Issue

2

fYear

2004

Firstpage

269

Lastpage

274

Abstract

We integrate the tangent method into a statistical framework for classification analytically and practically. The resulting consistent framework for adaptation allows us to efficiently estimate the tangent vectors representing the variability. The framework improves classification results on two real-world pattern recognition tasks from the domains handwritten character recognition and automatic speech recognition.

Keywords

estimation theory; handwritten character recognition; image recognition; speech recognition; statistical analysis; vectors; adaptation; automatic speech recognition; domains handwritten character recognition; real world pattern recognition; statistical framework; statistical pattern recognition; tangent method; tangent vectors estimation; Automatic speech recognition; Bayesian methods; Character recognition; Handwriting recognition; Kernel; Maximum likelihood estimation; Pattern recognition; Probability density function; Training data; Vectors; Algorithms; Artificial Intelligence; Automatic Data Processing; Cluster Analysis; Feedback; Image Enhancement; Image Interpretation, Computer-Assisted; Information Storage and Retrieval; Models, Statistical; Natural Language Processing; Numerical Analysis, Computer-Assisted; Pattern Recognition, Automated; Reading; Reproducibility of Results; Sensitivity and Specificity; Signal Processing, Computer-Assisted; Speech Perception; Subtraction Technique;

fLanguage

English

Journal_Title

Pattern Analysis and Machine Intelligence, IEEE Transactions on

Publisher

ieee

ISSN

0162-8828

Type

jour

DOI

10.1109/TPAMI.2004.1262198

Filename

1262198