13th Annual Conference of the International Speech Communication Association

Portland, OR, USA
September 9-13, 2012

Feature Selection for Speaker Traits

Jouni Pohjalainen (1), Serdar Kadioglu (2), Okko Räsänen (1)

(1) Department of Signal Processing and Acoustics, Aalto University, Espoo, Finland
(2) Department of Computer Science, Brown University, Providence, RI, USA

This study focuses on handling high-dimensional classification problems by means of feature selection. The data sets used are provided by the organizers of the Interspeech 2012 Speaker Trait Challenge. A combination of two feature selection approaches gives results that approach or exceed the challenge baselines using a k-nearest-neighbor classifier. One of the feature selection methods is based on covering the data set with correct unsupervised or supervised classifications according to individual features. The other selection method applies a measure of statistical dependence between discretized features and class labels.

Index Terms: pattern recognition, feature selection, high-dimensional data, speaker characteristics

Full Paper

Bibliographic reference.  Pohjalainen, Jouni / Kadioglu, Serdar / Räsänen, Okko (2012): "Feature selection for speaker traits", In INTERSPEECH-2012, 270-273.