Abstract:

In a world like today, where days start and end with smart voice assistants, speech emotion recognition plays a major role. The development in science is incredible as it is making the human life easy and simple. Speech emotion recognition is the concept in which a machine can be able to sense the emotions of human. Instead of using hybrid classifiers, we chose only one classifier, that is, MLP (Multi-Layer Perceptron). With the help of MLP we built a machine learning model that gives above 80% accuracy. We developed an application in such a way that it takes instant inputs from end users and predicts emotion based on their voice note. Also, the sound wave of that voice note can also be seen.

Description:

Supervised Learning, MLP-Classifier, MFCC, Model, Feature Extraction