Deception detection in speech using bark band and perceptually significant energy features

Author

Sanaullah, Muhammad ; Gopalan, Kartik

Author_Institution

Electr. & Comput. Eng. Dept., Purdue Univ. Calumet, Calumet, IN, USA

fYear

2013

fDate

4-7 Aug. 2013

Firstpage

1212

Lastpage

1215

Abstract

This paper presents the initial results of analysis of nonlinear spectral features for classifying truthful and deceptive speech. These features are derived on a Bark scale based on the psychoacoustic masking property of human speech perception. Truthful and deceptive speeches are established a posteriori by a male speaker under jeopardy. Test results using significant energy features at Bark bands and a neural network have a potential to show delicate variations between truthful and deceptive utterances.

Keywords

acoustic signal processing; neural nets; signal classification; spectral analysis; speech intelligibility; Bark bands; Bark scale; deception detection; deceptive speech classification; deceptive utterances; energy features; human speech perception; jeopardy; male speaker; neural network; nonlinear spectral features; psychoacoustic masking property; truthful speech classification; truthful utterances; Bark Energy; Bark Scale; Deceptive Speech; Neural Network; Psychoacoustic; Significant Energy;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems (MWSCAS), 2013 IEEE 56th International Midwest Symposium on

Conference_Location

Columbus, OH

ISSN

1548-3746

Type

conf

DOI

10.1109/MWSCAS.2013.6674872

Filename

6674872