|
Deep Learning Methodology for Recognition of Emotions using Acoustic features
|
|
| Author:
|
VENKATESH MANNEM, SWARNA KUCHIBHOTLA
|
| Abstract:
|
Speech is the natural way of communication through which one can express emotions clearly. The same scenario is extended to computer applications Viz Speech emotion recognition (SER) system. In this work, we have used several acoustic features from speech signals for determining the emotional state. The process of designing a SER system is one of the most challenges in current research areas. A Deep Learning Multi-Layer Perceptron (MLP) algorithm is used for classification of various emotions. A comparative analysis has been made on accuracies of MLP using different combination of voice features like MelFrequency Cepstral Coefficients (MFCC), Chroma and Mel features. We obtained an improvement in the accuracies of MLP classifier with MFCC features.
|
| Keyword:
|
Deep Learning, Speech Emotion Recognition, Multi Layer Perceptron, RAVDESS dataset, Multi Dimensional Dataset, Classification, Attribute Analysis.
|
| EOI:
|
-
|
| DOI:
|
https://doi.org/10.31838/ijpr/2020.12.04.462
|
| Download:
|
Request For Article
|
|
|