Abstract

Speech Emotion Recognition (SER) system is an emerging field in the domain of human-computer interaction that aims to identify and analyze the emotional state of a speaker based on their speech signals. This project report presents the development of a Speech Emotion Recognition System that incorporates a user-friendly interface using HTML, CSS, and JavaScript, along with Django as the backend framework.

The system allows users to record their speech or import an audio file and predict their emotions accurately using a pre-trained model. The project demonstrates the potential of SER technology in various applications, including sentiment analysis, virtual assistants, customer service, and mental health monitoring

CHAPTER ONE

LITERATURE OVERVIEW

1. Overview

1.1 Purpose

The purpose of my project on “Speech Emotion Recognition System” is to design and develop a system that can accurately recognize and predict the emotional state of a speaker based on their speech signals. The project aims to provide a user-friendly interface where users can either record their voice or import an audio file to analyze and determine their emotional state.

The primary purposes of your project can be summarized as follows:

Emotion Recognition: The main objective of my project is to accurately recognize and classify the emotions expressed in speech signals. By leveraging machine learning and signal processing techniques, I aim to develop a system that can effectively analyze speech features and predict emotions such as happiness, sadness, anger, etc.

User Interaction: My project focuses on providing an intuitive and user-friendly interface for users to interact with the system. Users should be able to record their voice or upload an audio file through the web interface, and the system should provide real-time or near-real-time emotion predictions.

Practical Application: My project aims to demonstrate the practical application of Speech Emotion Recognition in various domains. This could include areas such as human-computer interaction, sentiment analysis, virtual assistants, or even mental health monitoring. By showcasing the potential use cases of the system, you aim to highlight the relevance and importance of emotion recognition technology.

Integration of Technologies: My project involves integrating multiple technologies and frameworks, including HTML, CSS, JavaScript for the frontend interface, Django for the backend development, and machine learning algorithms for emotion recognition. The purpose is to demonstrate my ability to combine different tools and techniques to create a functional and robust system.

Evaluation and Performance: My project includes evaluating the performance of the Speech Emotion Recognition system. This involves assessing the accuracy, precision, recall, and other evaluation metrics to measure the effectiveness of the developed model. The purpose is to provide insights into the system’s performance and identify areas for potential improvement.

Contribution to the Field: My  project contributes to the existing research and knowledge in the field of Speech Emotion Recognition. By documenting my methodology, experimental setup, and results, I aim to share your findings with the research community, potentially advancing the state-of-the-art in emotion recognition systems

1.1.1 Problem Statement

Traditional methods of speech emotion detection, such as using handcrafted features and statistical models, have limited accuracy and require extensive domain knowledge. Deep learning models have shown promising results in recent years, but there is still a need for optimized models that can accurately classify emotions in real-time.

Therefore, the proposed project aims to address this problem by developing a deep learning model for speech emotion detection that can accurately classify emotions in real-time, with potential applications in mental health monitoring, speech therapy, human-computer interaction, and marketing.

1.1.2 Study of the existing system

The existing  systems often rely on handcrafted features and conventional machine learning algorithms like Support Vector Machines (SVM) or Gaussian Mixture Models (GMM) eg EmoVoice. Which has the following disadvantages.

Limited Accuracy: Handcrafted features may not capture all relevant information for emotion recognition, leading to lower accuracy.

Time-Consuming Feature Engineering: Extracting meaningful features manually is time-consuming and requires domain expertise.

Lack of Real-time Processing: Conventional algorithms may not be optimized for real-time processing, causing delays in emotion predictions.

Check out: Computer Engineering Project Topics with Materials