Binary-Multi-CNN-CIFAR10

CNN Model for Binary & Multi-Class Classification using CIFAR-10 Dataset

Overview

This project showcases an image classification application built using TensorFlow and convolutional neural networks (CNNs) to classify images from the CIFAR-10 dataset. It includes two models:

• Binary Classification Model: Classifies images into two categories.
• Multi-Class Classification Model: Classifies images into ten different classes (e.g., dog, cat, horse).

Key Features

• Utilizes the CIFAR-10 dataset for training and testing.
• Implements two CNN models for binary and multi-class classification.
• Integrated into a Flask web application for easy image upload and classification.
• Potential for further enhancements like real-time classification and improved user interface.

Project Structure

Binary-Multi-CNN-CIFAR10/
│
├── app.py                      # Flask application for model deployment
├── model_binary.h5             # Pre-trained binary classification model
├── model_multiclass.h5         # Pre-trained multi-class classification model
├── static/                     # Static files (e.g., CSS, images)
│   └── styles.css              # Stylesheet for the web application
├── templates/                  # HTML templates for Flask
│   ├── index.html              # Homepage for image upload
│   └── result.html             # Result page for displaying predictions
├── README.md                   # Project documentation (this file)
└── requirements.txt            # Required packages

Getting Started

Prerequisites

Ensure you have Python installed. You can download it from python.org.

Install Required Packages

Install Flask and TensorFlow using the following command:

pip install Flask tensorflow

Organize Files

Place the provided code in the appropriate files and directories as described in the project structure.

Load Models

Ensure the paths to your trained models are correctly specified in app.py:

model_binary = load_model('model_binary.h5')
model_multiclass = load_model('model_multiclass.h5')

Running the Application

1. Navigate to the project directory:

   cd Binary-Multi-CNN-CIFAR10

2. Run the Flask application:

   python app.py

3. Open your browser and go to http://127.0.0.1:5000/ to access the application.

Models Overview

Data Loading and Preprocessing

• The CIFAR-10 dataset is loaded and normalized to a range between 0 and 1 for efficient model training.

Multi-Class Classification Model

• A CNN model is constructed to classify images into ten different classes.

Binary Classification Model

• A separate CNN model is built to classify images into two categories, demonstrating the adaptability of CNNs for binary tasks.

Future Enhancements

This project serves as a foundation for further development and enhancements:

1. Flask Web Application: Enhance the Flask app to allow users to upload images and receive classification predictions.
2. Improved User Interface: Utilize Bootstrap and CSS to create an engaging frontend.
3. Additional Classes: Expand the model to recognize a wider variety of objects.
4. Real-Time Classification: Implement real-time image classification using a device's camera.
5. Deployment to Production: Deploy the web application to a public server.
6. User Authentication: Add user-specific dashboards for managing classification history.
7. Model Fine-Tuning: Experiment with different architectures and hyperparameters to improve accuracy.

Usage

Upload an image through the web interface to get predictions from the binary and multi-class models.

Screenshots

Home Page:

Prediction Results:

Credits

This project is created and maintained by [Abhishek Shah].

License

This project is licensed under the MIT License.