Enhancing Authorship Attribution with Synthetic Paintings
This work tackles the problem of limited data for artwork authentication, which is a challenge for art historians and conservators.
This study addresses the scarcity of training data for attributing authorship to paintings by using synthetic images generated via DreamBooth fine-tuning of Stable Diffusion. The hybrid approach, combining real and synthetic data, resulted in improved ROC-AUC and accuracy compared to using only real paintings.
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.