Enhancing Authorship Attribution with Synthetic Paintings

TL;DR

This paper explores using synthetic images generated by fine-tuned diffusion models to improve authorship attribution accuracy for paintings, addressing data scarcity issues in art authentication.

Contribution

It introduces a hybrid approach combining real and synthetic images to enhance classification performance in artwork authorship attribution.

Findings

Synthetic images improve ROC-AUC and accuracy

Hybrid data approach outperforms real-only models

Generative-discriminative integration benefits art authentication

Abstract

Attributing authorship to paintings is a historically complex task, and one of its main challenges is the limited availability of real artworks for training computational models. This study investigates whether synthetic images, generated through DreamBooth fine-tuning of Stable Diffusion, can improve the performance of classification models in this context. We propose a hybrid approach that combines real and synthetic data to enhance model accuracy and generalization across similar artistic styles. Experimental results show that adding synthetic images leads to higher ROC-AUC and accuracy compared to using only real paintings. By integrating generative and discriminative methods, this work contributes to the development of computer vision techniques for artwork authentication in data-scarce scenarios.