Emotion detection using relative amplitude-based features through speech

Automatic speech recognition analysis has been an active part in computer science for more than two decades. In general, to detect an emotion, long continuous signal is needed. Relative amplitude reduces bias of glottal mutation of speech wave amplitude and obtains a normalized measure without conce...

Full description

Saved in:
Bibliographic Details
Main Authors: Mohan Kudiri, Krishna, Md Said, A., Nayan, M. Y.
Format: Conference Proceeding
Language:English
Subjects:
Artificial neural networks
Benchmark testing
Computational modeling
emotion detection
Hidden Markov models
Kernel
relative bin frequency features
relative sub-image based features
support vector machine
Support vector machines
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Automatic speech recognition analysis has been an active part in computer science for more than two decades. In general, to detect an emotion, long continuous signal is needed. Relative amplitude reduces bias of glottal mutation of speech wave amplitude and obtains a normalized measure without concern of information from being distinct in feature. Nonverbal communication plays crucial role in human-human or human-machine interpersonal relationships. In this paper, we propose the use of relative bin frequency coefficients for speech signal segmentation. Here, the support vector machine classifier is used to implement automatic emotion detection system.
DOI:10.1109/ICCISci.2012.6297301