Description

In recent years advanced multi spectral and X-ray imaging techniques have allowed art historians and preservation experts to find hidden images beneath famous paintings. These hidden images can have different origins. In some cases they are older paintings or sketches on a canvas that the artist decided to reuse for another painting. Canvases were often reused to save money. In other cases modifications to the original painting are added to conceal controversial details, or just because the artist changed their mind. We propose to use artificial intelligence in combination with high quality multi spectral data to simplify, automatize, and improve the process of analyzing a painting, identify its properties and find hidden images.

Duration

Total project length: 175 hours

Task ideas

• Extract useful information from multi spectral images of paintings, pigment type, damage, previous restorations.
• Find if the painting contains a hidden image that has been painted over.

Expected results

• Prepare a data set of multi spectral images of paintings.
• Train a computer vision model to identify painting properties
• Train a computer vision model to identify images hidden beneath paintings
• (If time permits) train an AI model to retrieve the hidden image.

Requirements

• Python
• PyTorch/TensorFlow
• Experience with computer vision

Project difficulty level

Medium

Mentors

• Emanuele Usai (University of Alabama)
• Sergei Gleyzer (University of Alabama)

Test

Please use this link to access the test for this project.

Please DO NOT contact mentors directly by email. Instead, please email human-ai@cern.ch with Project Title and include your CV and test results. The mentors will then get in touch with you.

Corresponding Project

• ArtExtract

Participating Organizations

• Alabama

---