/M.U.R.A.L

M.U.R.A.L stands for Machine-aided Understanding and Recognition of Artistic Legacies. It is an application made by me to identify if a piece of art is made by human artist or by a generative AI algorithm.This helps the artist gain the credit they deserve for their hard work and helps to keep the "Human Essence of Art" Intact in the era of AI.

Primary LanguagePureBasicMIT LicenseMIT

M.U.R.A.L : Machine-aided Understanding and Recognition of Artistic Legacies

Introduction

M.U.R.A.L stands for Machine-aided Understanding and Recognition of Artistic Legacies. It is an application designed to identify artworks created by human artists and those generated by AI algorithms. This initiative aims to ensure proper credit for human artists and preserve the "Human Essence of Art" in the age of AI.

  • Important details:
    • Theme is custom made and packed with application as a .json file in details
    • The model folder contains the CNN model that i have trained myself using the keras and tensorflow.The linke to this ML code is given below
    • The application is made using the customtkinter module by python and has a frontend (Mainpage.py or mainpage2.py) and a backend (Backend.py) code.

Functionality

  • Artwork Identification: M.U.R.A.L employs advanced algorithms to identify and categorize artworks.
  • Human vs AI Recognition: The application determines whether a piece of art is crafted by a human artist or generated by AI.
  • Credit Attribution: By distinguishing between human and AI-generated art, M.U.R.A.L aids in giving due credit to human creators.

Technological Stack

  • Python
  • Tkinter (GUI Framework)
  • TensorFlow and Keras (Machine Learning)
  • PIL (Python Imaging Library)
  • tqdm (Dynamic modules loading)
  • CustomTkinter (Custom UI Elements)
  • Numpy (Accelerated vector computing)
  • Backend (Custom backend module)

Application Components

  • About Frame: Provides information about the application.
  • Image Preview Frame: Allows users to load, clear, and preview images.
  • Prediction Frame: Initiates the process of predicting whether an image is human-generated or AI-generated, displaying results with a progress bar.
  • Info Frame: Offers additional information based on the prediction result.

Images of application

  • Following are the screenshots of my application
Mainpage 1 Mainpage 2

How to Use

  1. Launch the application using mainpage.py.
  2. Explore the "About," "Image Preview," and "Info" sections.
  3. Load an image, preview it, and initiate predictions to distinguish between human and AI-generated art.

Building and Running

  1. Install the required Python modules using the provided prerequisite.py.
  • run file:
    • path/prerequisite.py
  1. Execute the application:
  • Run either the python script
    • path/mainpage.py
  • Run executable if available

License

This project is not licensed yet.

About Author

This application is made entirely from scratch, including the building of the neural network and the application, by Malhar Girgaonkar. I have a deep appreciation for art in all its shapes and forms. The current trend of AI gaining popularity in creative fields, which is at the forefront of what defines humans, has caused many artists to lose their mental peace due to job security and replacement fears by the more powerful, productive, and efficient AI.

I believe that certain things cannot be replicated by AI, as art is something that comes from the soul and not just blatant learning from other artists, which I firmly believe is a form of artistic theft. As an artist and a student of computer science, I considered this project as my personal contribution to help art and artists get the recognition they deserve.

Contact

Contributing

Contributions to M.U.R.A.L are welcome. If you encounter issues or have ideas for improvement, please open an issue or submit a pull request.