Deep Ensemble Art Style Recognition

TL;DR

This paper advances art style recognition by evaluating multiple deep neural networks, introducing a stacking ensemble method, and achieving state-of-the-art accuracy on large art datasets, with insights into data preprocessing and style impact.

Contribution

It compares eight deep architectures on art style recognition, introduces a novel ensemble approach, and provides a comprehensive analysis of data and style effects.

Findings

Achieved state-of-the-art accuracy of 68.55% on WikiArt dataset.

Demonstrated the effectiveness of ensemble stacking with multiple models.

Identified the influence of data preprocessing and art styles on model performance.

Abstract

The massive digitization of artworks during the last decades created the need for categorization, analysis, and management of huge amounts of data related to abstract concepts, highlighting a challenging problem in the field of computer science. The rapid progress of artificial intelligence and neural networks has provided tools and technologies that seem worthy of the challenge. Recognition of various art features in artworks has gained attention in the deep learning society. In this paper, we are concerned with the problem of art style recognition using deep networks. We compare the performance of 8 different deep architectures (VGG16, VGG19, ResNet50, ResNet152, Inception-V3, DenseNet121, DenseNet201 and Inception-ResNet-V2), on two different art datasets, including 3 architectures that have never been used on this task before, leading to state-of-the-art performance. We study the effect of data preprocessing prior to applying a deep learning model. We introduce a stacking ensemble method combining the results of first-stage classifiers through a meta-classifier, with the innovation of a versatile approach based on multiple models that extract and recognize different characteristics of the input, creating a more consistent model compared to existing works and achieving state-of-the-art accuracy on the largest art dataset available (WikiArt - 68,55%). We also discuss the impact of the data and art styles themselves on the performance of our models forming a manifold perspective on the problem.