Facial Keypoints Detection Using CNN

This repository contains a deep learning model that predicts facial landmarks using Convolutional Neural Networks (CNN). Facial landmarks are key points on a person's face, such as the corners of the eyes, nose, and mouth. This project uses a CNN architecture to detect and predict these landmarks, which can be useful in various applications, including facial recognition, emotion analysis, and augmented reality.

Model Architecture

Sequential([
    Conv2D(16, 3, activation='relu', input_shape=(96, 96, 1)),
    MaxPooling2D(2),
    
    Conv2D(32, 3, activation='relu'),
    MaxPooling2D(2),
    
    Conv2D(64, 3, activation='relu'),
    MaxPooling2D(2),
    
    Conv2D(128, 3, activation='relu'),
    MaxPooling2D(2),
    
    Flatten(),
    Dense(512, activation='relu'),
    Dropout(0.2),
    
    Dense(512, activation='relu'),
    Dropout(0.2),
    
    Dense(30, activation='tanh')
])

Training Results

• Final Training Loss (MSE): 9.37e-04
• Final Training MAE: 0.0230
• Validation Loss (MSE): 0.0393
• Validation MAE: 0.0728
• Epochs: 150
• Batch Size: 64
• Validation Split: 20%

Key Features

• Input: 96x96 grayscale images
• Data augmentation with horizontal flipping
• Custom data loader with normalization
• Mean Squared Error (MSE) loss with Adam optimizer
• Model checkpointing for best validation performance

Requirements

• Python 3.x
• TensorFlow/Keras
• NumPy
• Pandas
• Matplotlib (for visualization)

Sample Input

Sample Output