Automated Artist Identification Using Deep Learning and Transfer Learning Techniques

Abstract

Traditional methods of art analysis rely heavily on human expertise, which can be subjective and limited by an individual’s familiarity with art history.This paper introduces an innovation of art style recognition using deep learning and transfer learning techniques. We use the pre-trained ResNet50 architecture to identify artists based on their work. The study aims to show how convolutional neural networks (CNNs) can be applied to complex image recognition tasks, such as identifying artists from their paintings. Our model is trained on a large dataset of digitized artworks with astonishing accuracy, demonstrating the effectiveness of transfer learning in dealing with the high variability of artistic styles.