Sentiment Analysis-Emotion Recognition

Document Type : Primary Research paper

Authors

1 B. Tech (Scholar), School of Computer Science & Engineering, Sandip University, Nashik, India

2 Assistant Professor, School of Computer Science & Engineering, Sandip University, Nashik, India

3 Professor, School of Computer Science & Engineering, Sandip University, Nashik, India

Abstract

One of the key elements of emotional computing in human-computer interaction is voice communication. The aim of this research is to build and create a speech-based emotion reaction (SER) prediction system in which various emotions are identified using CNN classifiers. Mel-frequency cepstral (MFCC) is an extracted spectral characteristic. The suggested method is developed using the Librosa module in Python, and its effectiveness is evaluated using samples from the Database of Emotional Speech and Song (RAVDESS), which allows users to distinguish between emotions including happiness, surprise, anger, neutrality, sadness, fear, etc. To find the most important feature subset, feature selection (FS) was used. Results indicate that utilizing CNN gives good performance

Keywords

Int. J. of Aquatic Science
Volume 14, Issue 1
2023
Pages 381-390

Files

Share

How to cite

Statistics