FASHION MNIST FLASK APPLICATION

Directory Structure

/your_app/ ├ app.py # Flask application entry point ├ static/ │ ├ css/ └ style.css │ ├ data/ └ Fmnist_dataset.ipynb # Jupyter notebook for downloading and processing the dataset │ └ js/ └ mnist.js ├ templates/ │ └ index.html # HTML template for the UI ├ models/ │ └ F_mnist_model.pth # Pre-trained model file └ modules/ └ data_processing.py # Image processing logic └ model.py

Installation

1. Create a virtual environment (optional but recommended):

   python -m venv venv (if already have can use your own venv)
   

2. Change the user preference for the PowerShell script execution policy Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope Process

3. Activate venv

    source venv/bin/activate
   # On Windows use `venv\Scripts\activate`
   

4. Install dependencies:

   pip install -r requirements.txt

5. Install Jupyter (If Not Installed):

   pip install jupyter

6. **Download the Fashion MNIST dataset and save images **: -Navigate to the static\data folder and open fmnist_dataset.ipynb -Run the fmnist_dataset.ipynb notebook code 1 and 4 to download the dataset and save a few example images. The Fmnist_images folder will contain subfolders for each class, and inside each subfolder, 3 images corresponding to that class will be saved. This is necessary for testing the app. -Once the F_mnist_dataset.ipynb file is open, you can run the notebook step by step. To run a cell, click on it and press Shift + Enter.

Usage

1. Run the Flask application:

   python app.py

2. Access the application:

   • http://127.0.0.1:5000 will directly go to the main web browser.

3. Image Upload:

   • Users can upload an image of fashion items (such as T-shirts, trousers, or shoes) via the index.html page.
   • You can used the save images inside the Fmnist_images folder to test the website.
   • Once the image is uploaded and processed, the predicted class label is displayed on the page.

Error Handling

• No file uploaded: If no file is selected during the image upload, the app will return an error message.
• Invalid file type: Only image files are accepted. If a non-image file is uploaded, an error will be shown.
• Prediction errors: If an error occurs during the prediction process (e.g., model loading issue), the app will return a failure message with an error description.