Art Style Classifier with TensorFlow and Flask

Overview

This project uses TensorFlow and Keras to build a Convolutional Neural Network (CNN) for classifying two different styles of art: Japanese and Rococo. The implementation includes data preparation, model training, and a Flask application for making predictions.

Data Source

The dataset used in this project is sourced from WikiArt - Art Movements/Styles.

Project Structure

• data_preparation.ipynb: Jupyter Notebook for reading images and creating train, test, and validate datasets.
• implementation.ipynb: Jupyter Notebook for reading and augmenting images, building and training the CNN model, and running validation.
• app/:
  • app.py: Flask application for predicting the art style using the trained model.
  • static/: Folder for storing static files (images in this case).
  • templates/: HTML templates for the Flask application.

Usage

run the flask app using python app.py Make sure you have all the dependancies installed