📷 CNN Visualization for Glasses Classification

🧠 A Convolutional Neural Network (CNN) project to classify images of glasses vs. no-glasses with a real-time visualization interface created without using any libraries related to neural networks.

---

✨ Features

📊 Interactive GUI

A Tkinter-based GUI that showcases:

• The CNN Structure:
  • Input, Convolutional, Pooling, and Fully Connected layers.
• Real-Time Training Metrics:
  • Loss and Accuracy graphs that update dynamically.
• Dynamic Weight Visualization:
  • Connections are color-coded to represent weight strength after each epoch.

---

🛠 Comprehensive CNN Architecture

1. Convolutional Layer: Extracts spatial features from the input using filters.

2. MaxPooling Layer: Reduces the spatial dimensions while retaining the most important features.

3. Fully Connected Layer: Maps high-level features to class probabilities.

---

📂 Dataset

A collection of labeled images, classified into two categories:

• Glasses
• No-Glasses

---

📌 Customizable

• Hyperparameters:
  • Learning rate
  • Number of epochs
• Network Layers:
  • Easily modify the number of filters, pooling size, and fully connected neurons.

---

🚀 Getting Started

1️⃣ Prerequisites

• Required libraries: numpy, scipy, pillow, matplotlib, tkinter, scikit-learn