https://github.com/sabin74/image_classification_cifar10

This project demonstrates how to build a deep learning image classifier using the CIFAR-10 dataset. Two approaches are implemented: A custom Convolutional Neural Network (CNN) & A transfer learning model using VGG16
https://github.com/sabin74/image_classification_cifar10

batch-normalization cifar-10-dataset cnn-a data-augmentation evaluation image-classification tensorflow transfer-learning vgg16-tensorflow

Last synced: 6 months ago
JSON representation

This project demonstrates how to build a deep learning image classifier using the CIFAR-10 dataset. Two approaches are implemented: A custom Convolutional Neural Network (CNN) & A transfer learning model using VGG16

• Host: GitHub
• URL: https://github.com/sabin74/image_classification_cifar10
• Owner: sabin74
• Created: 2025-06-14T06:45:47.000Z (9 months ago)
• Default Branch: main
• Last Pushed: 2025-06-14T06:49:47.000Z (9 months ago)
• Last Synced: 2025-06-14T07:43:41.069Z (9 months ago)
• Topics: batch-normalization, cifar-10-dataset, cnn-a, data-augmentation, evaluation, image-classification, tensorflow, transfer-learning, vgg16-tensorflow
• Homepage: https://www.cs.toronto.edu/~kriz/cifar.html
• Size: 0 Bytes
• Stars: 0
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# 🧠 CIFAR-10 Image Classification using CNN & Transfer Learning

This project demonstrates how to build a deep learning image classifier using the CIFAR-10 dataset. Two approaches are implemented:

1. A custom Convolutional Neural Network (CNN)

2. A transfer learning model using VGG16

---

## 📊 Dataset Overview

- **Dataset:** CIFAR-10

- **Classes (10):** airplane, automobile, bird, cat, deer, dog, frog, horse, ship, truck

- **Image Shape:** (32, 32, 3)

- **Train/Test Split:** 50,000/10,000

---

## 🚀 Model Architectures

### Custom CNN

- 3 Convolutional blocks with MaxPooling

- Batch Normalization and Dropout

- Dense layers for final classification

### Transfer Learning (VGG16)

- `VGG16` with `include_top=False`

- Input images resized to (224x224)

- GlobalAveragePooling + Dense layers

---

## 🛠️ Key Techniques

- CNN Architecture

- Data Augmentation

- Regularization (Dropout, L2)

- Batch Normalization

- Transfer Learning

- Callbacks (EarlyStopping, ModelCheckpoint)

- Evaluation (Confusion Matrix, Classification Report)