Paper Info
Title
Robust Angry Speech Detection Employing A Teo-Based Discriminative Classifier Combination
Abstract
This study proposes an effective angry speech detection approach employing the TEO-based feature extraction. Decorrelation processing is applied to the TEO-based feature to increase model training ability by decreasing the correlation between feature elements and vector size. Minimum classification error training is employed to increase the discrimination between the angry speech model and other stressed speech models. Combination with the conventional Mel frequency cepstral coefficients (MFCC) is also employed to leverage the effectiveness of MFCC to characterize the spectral envelope of speech signals. Experimental results over the SUSAS corpus demonstrate the proposed angry speech detection scheme is effective at increasing detection accuracy on an open-speaker and open-vocabulary task. An improvement of up to 7.78% in classification accuracy is obtained by combination of the proposed methods including decorrelation of TEO-based feature vector, discriminative training, and classifier combination.
Year
Venue
Keywords
2009
INTERSPEECH 2009: 10TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION 2009, VOLS 1-5
angry speech detection, TEO-based feature, discriminative training, classifier combination
Field
DocType
Citations 
Mel-frequency cepstrum,Feature vector,Decorrelation,Spectral envelope,Pattern recognition,Computer science,Voice activity detection,Feature extraction,Speech recognition,Artificial intelligence,Classifier (linguistics),Discriminative model
Conference
2
PageRank 
References 
Authors
0.47
9
2
Name
Order
Citations
PageRank
Wooil Kim112016.95
John H. L. Hansen23215365.75