Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Art Forgery Detection using Kolmogorov Arnold and Convolutional Neural Networks

Sandro Boccuzzo, Deborah Desirée Meyer, Ludovica Schaerf

TL;DR

This work uses a carefully compiled dataset of known artists forged by Beltracchi and a set of known works by the forger to train a multiclass image classification model based on EfficientNet, which shows a general agreement between the different models' predictions on artworks flagged as forgeries.

Abstract

Art authentication has historically established itself as a task requiring profound connoisseurship of one particular artist. Nevertheless, famous art forgers such as Wolfgang Beltracchi were able to deceive dozens of art experts. In recent years Artificial Intelligence algorithms have been successfully applied to various image processing tasks. In this work, we leverage the growing improvements in AI to present an art authentication framework for the identification of the forger Wolfgang Beltracchi. Differently from existing literature on AI-aided art authentication, we focus on a specialized model of a forger, rather than an artist, flipping the approach of traditional AI methods. We use a carefully compiled dataset of known artists forged by Beltracchi and a set of known works by the forger to train a multiclass image classification model based on EfficientNet. We compare the results with Kolmogorov Arnold Networks (KAN) which, to the best of our knowledge, have never been tested in the art domain. The results show a general agreement between the different models' predictions on artworks flagged as forgeries, which are then closely studied using visual analysis.

Art Forgery Detection using Kolmogorov Arnold and Convolutional Neural Networks

TL;DR

This work uses a carefully compiled dataset of known artists forged by Beltracchi and a set of known works by the forger to train a multiclass image classification model based on EfficientNet, which shows a general agreement between the different models' predictions on artworks flagged as forgeries.

Abstract

Art authentication has historically established itself as a task requiring profound connoisseurship of one particular artist. Nevertheless, famous art forgers such as Wolfgang Beltracchi were able to deceive dozens of art experts. In recent years Artificial Intelligence algorithms have been successfully applied to various image processing tasks. In this work, we leverage the growing improvements in AI to present an art authentication framework for the identification of the forger Wolfgang Beltracchi. Differently from existing literature on AI-aided art authentication, we focus on a specialized model of a forger, rather than an artist, flipping the approach of traditional AI methods. We use a carefully compiled dataset of known artists forged by Beltracchi and a set of known works by the forger to train a multiclass image classification model based on EfficientNet. We compare the results with Kolmogorov Arnold Networks (KAN) which, to the best of our knowledge, have never been tested in the art domain. The results show a general agreement between the different models' predictions on artworks flagged as forgeries, which are then closely studied using visual analysis.
Paper Structure (20 sections, 3 figures, 6 tables)

This paper contains 20 sections, 3 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Methodology
  4. Pre-processing
  5. Models
  6. Efficientnets
  7. Kolmogorov Arnold Network (KAN)
  8. Data
  9. Evaluation
  10. Entropy comparison
  11. Blurring comparison
  12. Model size comparison
  13. Comparison with KAN
  14. Flagging questionable patches
  15. Results
  16. ...and 5 more sections

Figures (3)

  • Figure 1: Validation accuracy in % over EfficientNet $B0$ - $B6$ using dataset $T38$ - $T47$ and entropy > 2.5.
  • Figure 2: Image of the painting "Still Life with Pears and Indian Bowl" attributed to André Derain that where attributed to Wolfgang Beltracchi by 4 of our 7 Models trained with $T38$. Image in the public domain.
  • Figure 3: Image of the painting "Still Life with Pears and Indian Bowl" attributed to André Derain that where attributed to Wolfgang Beltracchi by KAN Model (Width 120/84/12, Grid = 5, k = 3) trained with $T38$. Image in the public domain.