Combining Prosodic Lexical and Cepstral Systems for Deceptive Speech Detection

A system combination approach using different models and features for deception detection is proposed, resulting in improved accuracy over the individual systems.

Abstract

We report on machine learning experiments to distinguish deceptive from non-deceptive speech in the Columbia-SRI-Colorado (CSC) corpus. Specifically, we propose a system combination approach using different models and features for deception detection. Scores from an SVM system based on prosodic/lexical features are combined with scores from a Gaussian mixture model system based on acoustic features, resulting in improved accuracy over the individual systems. Finally, we compare results from the prosodic-only SVM system using features derived either from recognized words or from human transcriptions.