Title of Project: VOICE SENTIMENTAL ANALYSIS

Team Members:

1. RAGHAV GOEL (2100290100125)
2. PRIYANSHU VISHWAKARMA (2100290100122)
3. HARSHIT SANGAL (2100290100071)
4. JITESH KUMAR (2100290100076)

Steps to Execute: Voice Sentimental Analysis

1. Setup Environment Open the project in Anaconda, Jupyter Lab, or Jupyter Notebook.

2. Install Dependencies Run the following to install required libraries:

   pip install pandas numpy librosa tensorflow scikit-learn matplotlib sounddevice scipy

3. Download Dataset Place the CSV dataset in the same folder as the notebook to train or re-train the model.

4. Run the Notebook Execute all cells in the provided Jupyter notebook to:

   i. Load and preprocess data

   ii. Extract features

   iii. Train the model

   iv. Save/load the trained model

5. Record Your Voice Run the audio recorder script or its corresponding cell to record your voice (e.g., 3–5 seconds). It saves the audio as a .wav file.

6. Predict Sentiment Run the predictor function to analyze the recorded voice and output the emotion (e.g., neutral, calm, happy, sad, angry, fearful, disgust, surprised).

Checklist:

1. Final Project Report
2. Certificate VII Semester (Dated: December 2024).
3. Certificate VIII Semester (Dated: May 2025).
4. Synopsis
5. Final Presentation
6. Source Code
7. Database dump (.sql file)
8. If a web project, then a Docker file for deployment
9. README (This file)