Exploring the benefits of discretization of acoustic features for speech emotion recognition

被引：0

作者：

Vogt, Thurid ^{[1
]}

Andre, Elisabeth ^{[1
]}

机构：

[1] Univ Augsburg, D-8900 Augsburg, Germany

来源：

INTERSPEECH 2009: 10TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION 2009, VOLS 1-5 | 2009年

关键词：

speech emotion recognition; discretization of features;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

We present a contribution to the Open Performance sub-challenge of the INTERSPEECH 2009 Emotion Challenge. We evaluate the feature extraction and classifier of EmoVoice, our framework for real-time emotion recognition from voice on the challenge database and achieve competitive results. Furthermore, we explore the benefits of discretizing numeric acoustic features and find it beneficial in a multi-class task.

引用

页码：348 / 351

页数：4

共 15 条

[1]

[Anonymous], P WORKSH PERC INT TE

[2]

[Anonymous], 2001, P EUR

[3]

[Anonymous], INTERSPEECH

[4]

[Anonymous], Data Mining Practical Machine Learning Tools and Techniques with Java

[5]

[Anonymous], 13 INT JOINT C ARTIF

[6]

Boersma P., Praat: Doing Phonetics by Computer

[7]

Boersma P., 1993, Institute of Phonetic Sciences, University of Amsterdam, Proceedings 17 (1993) 97-110, P97

[8]

Casale S., 2008, 2008 Second IEEE International Conference on Semantic Computing (ICSC), P158, DOI 10.1109/ICSC.2008.43

[9]

FINK G, 1999, LECT NOTES ARTIF INT, V1962, P229

[10]

Gilroy SW., 2008, P 16 ACM INT C MULTI, P945

← 1 2 →