https://github.com/aronno1920/fashion-classifier-mnist
This project is a full-stack web application that allows users to upload an image of a fashion item (e.g., shirt, sneaker, bag) and classifies it using a Neural Network (NN) and Convolutional Neural Network (CNN)
https://github.com/aronno1920/fashion-classifier-mnist
cnn convolutional-neural-networks deep-learning fashion-classifier flask mnist nn nuralnetwork python
Last synced: about 2 months ago
JSON representation
This project is a full-stack web application that allows users to upload an image of a fashion item (e.g., shirt, sneaker, bag) and classifies it using a Neural Network (NN) and Convolutional Neural Network (CNN)
- Host: GitHub
- URL: https://github.com/aronno1920/fashion-classifier-mnist
- Owner: Aronno1920
- Created: 2025-07-29T06:43:49.000Z (11 months ago)
- Default Branch: main
- Last Pushed: 2025-07-31T07:00:47.000Z (11 months ago)
- Last Synced: 2025-08-06T11:56:29.014Z (10 months ago)
- Topics: cnn, convolutional-neural-networks, deep-learning, fashion-classifier, flask, mnist, nn, nuralnetwork, python
- Language: Jupyter Notebook
- Homepage:
- Size: 10.9 MB
- Stars: 0
- Watchers: 0
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 🧥 Fashion Classifier (MNIST) – TensorFlow, Flask, and FastAPI
This project is a full-stack web application that allows users to upload an image of a fashion item (e.g., shirt, sneaker, bag) and classifies it using a Neural Network (NN) and Convolutional Neural Network (CNN) trained on the [Fashion Classifier](https://github.com/Aronno1920/Fashion-Classifier-MNIST) dataset.
The app features:
- 🔍 Real-time image classification
- ⚙️ NN and CNN model build separately
- 🧠 Robust preprocessing (resize, normalize, center the object)
- 📸 Live preview of uploaded image
- 💡 Displays top 3 predictions with confidence scores
- 🎨 Beautiful and responsive HTML/CSS interface
## 🚀 Live Demo
> _Want to see it in action?_
> 👉 [Coming Soon: Deployed on Render or Hugging Face Spaces]
## 🧠 Model Details
- **Architecture**: 3-layer CNN with BatchNorm, Dropout, and MaxPooling
- **Input shape**: 28×28 grayscale images
- **Training Dataset**: Fashion MNIST (60,000 training + 10,000 test images)
- **Accuracy**: ~91% on test data
---
## 🖼️ Supported Classes
``` ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat','Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot'] ```
---
## 🧪 Try It Locally
### 1️⃣ Clone the repo
```bash
git clone https://github.com/Aronno1920/Fashion-Classifier-MNIST.git
cd fashion-classifier-mnist
```
2️⃣ Install requirements
```bash
pip install -r requirements.txt
```
3️⃣ Run the app
```bash
python app.py
```
Open your browser: http://127.0.0.1:5000
📁 Project Structure
```Image-Classification-AI/
├── ImageSample/ # Folder for example images or datasets
├── __pycache__/ # Python cache files (auto-generated)
├── model/
│ ├── best_nn_model.keras # Saved trained NN model weights
│ └── best_cnn_model.keras # Saved trained CNN model weights
├── notebooks/
│ ├── fashion_mnist_selim_ahmed.py # Saved colab file as python
│ └── Fashion_MNIST_Selim_Ahmed.ipynb # Saved colab file
├── static/ # logo, style and other necessary
│ ├── logo.png
│ └── style.css
├── templates/ # Templates for web app or reports (if any)
│ ├── index.html
│ └── train.html
├── templates/
├── README.md # Project description and documentation
├── app.py # Main application script (e.g., for running or serving the model)
├── model_config.py
├── train_model_cnn.py # CNN model training class
└── train_model_nn.py # NN model training class
```
🛠 Tech Stack
```bash
Python
TensorFlow / Keras
Flask
PIL (Pillow)
HTML5 + CSS
JavaScript (Image preview)
SciPy (for center-of-mass preprocessing)
```
📸 Screenshots
✍️ Author
Selim Ahmed
📫 [Connect on LinkedIn](https://www.linkedin.com/in/aronno1920/)
🌐 [GitHub Profile](https://github.com/aronno1920)