This project focuses on image classification using the CIFAR-10 dataset. Two models, ANN and CNN, are trained on 32x32 color images spanning 10 different classes. The dataset undergoes preprocessing, after which the models are built, trained, and evaluated.
- 🔹 Artificial Neural Network (ANN) implementation
- 🔹 Convolutional Neural Network (CNN) implementation
- 🔹 Utilizes TensorFlow and Keras
- 🔹 CIFAR-10 dataset for multi-class classification
- 🔹 Model evaluation & visualization with Matplotlib
The CIFAR-10 dataset consists of 60,000 images (50,000 for training, 10,000 for testing) classified into:
🔹 Airplane
🔹 Automobile 🚗
🔹 Bird 🐦
🔹 Cat 🐱
🔹 Deer 🦌
🔹 Dog 🐶
🔹 Frog 🐸
🔹 Horse 🐴
🔹 Ship 🚢
🔹 Truck 🚛
An interactive GUI is included for testing the model with custom images.
- Upload an image for classification
- Get real-time predictions
- Compare ANN vs CNN performance
Ensure you have Python installed along with required libraries:
pip install tensorflow numpy matplotlib tkintergit clone https://github.com/yourusername/image-classification.git
cd image-classificationpython train_model.pypython gui.pyThe models are trained and evaluated based on accuracy and loss metrics.
| Model | Accuracy (Test) |
|---|---|
| ANN | 65% |
| CNN | 85% |
Feel free to contribute! Open an issue or submit a pull request.
This project is licensed under the MIT License.
For any queries, reach out to me at: your.email@example.com