Recognition of emotional state in Polish speech: comparison between human and automatic efficiency

Authors:
Piotr Staroniewicz
Affiliations:
Wroclaw University of Technology, Institute of Telecommunications, Teleinformatics and Acoustics, Wroclaw, Poland
Venue:
BioID_MultiComm'09 Proceedings of the 2009 joint COST 2101 and 2102 international conference on Biometric ID management and multimodal communication
Year:
2009

Citing 4
Cited 2

Emotional speech: towards a new generation of databases

Speech Communication - Special issue on speech and emotion
Vocal communication of emotion: a review of research paradigms

Speech Communication - Special issue on speech and emotion
Data Mining: Practical Machine Learning Tools and Techniques, Second Edition (Morgan Kaufmann Series in Data Management Systems)

Data Mining: Practical Machine Learning Tools and Techniques, Second Edition (Morgan Kaufmann Series in Data Management Systems)
Polish Emotional Speech Database --- Recording and Preliminary Validation

Cross-Modal Analysis of Speech, Gestures, Gaze and Facial Expressions

Automatic recognition of emotional state in Polish speech

Proceedings of the Third COST 2102 international training school conference on Toward autonomous, adaptive, and context-aware multimodal interfaces: theoretical and practical issues
Influence of speakers' emotional states on voice recognition scores

COST'10 Proceedings of the 2010 international conference on Analysis of Verbal and Nonverbal Communication and Enactment

Quantified Score

Hi-index	0.00

Visualization

Abstract

The paper presents the comparison of human (listeners test) and automatic (SVM classifier) speech emotion recognition. The database of Polish emotional speech used during tests includes recordings of six acted emotional states (anger, sadness, happiness, fear, disgust, surprise) and the neutral state of 13 amateur speakers (2118 utterances). The automatic classifier used the set of 31 attribute evaluated features, C-SVC algorithm with the Gaussian Radial Basis Function. The mean overall score for human recognition (57.25%) turned out to be lower than for automatic recognition (64.77%).