Emojize Ur Face!

Overview

"Emojize Ur Face!" is an innovative application that uses machine learning to analyze facial expressions in real-time and overlay corresponding emojis on the user's face. This project aims to enrich digital interactions by adding an expressive, fun twist.

How to Load and Run

Prerequisites

Ensure you have the following prerequisites installed:

• Python
• numpy
• opencv-python (cv2)
• joblib
• sklearn
• mediapipe

Steps

1. Download the Project: Clone or download the project repository to your local machine.
2. Obtain the Dataset: Download the FER2013 dataset from Kaggle and put them into the program directory.
3. Train the Model: Run train_model.py to train the emotion detection model.
4. Run the Application: Execute emojize.py to start the application. This script captures video from a webcam, analyzes facial expressions, and overlays emojis on the face in real-time.

Machine Learning Model and Strategy

The application employs a machine learning pipeline that includes Principal Component Analysis (PCA) for dimensionality reduction and a RandomForestClassifier for emotion detection. The workflow is as follows:

• Data Preparation: Images are converted to grayscale, resized, and flattened into vectors.
• PCA: Reduces the dimensionality of the image data, improving efficiency.
• Random Forest Classifier: An ensemble method used for classifying the emotion based on the processed image data.
• Model Evaluation: The model's performance is evaluated using accuracy as the metric.