No abstract

The increasing availability of large digitized fine art collections opens new research perspectives in the intersection of artificial intelligence and art history. Motivated by the successful performance of Convolutional Neural Networks (CNN) for a wide variety of computer vision tasks, in this paper we explore their applicability for art-related image classification tasks. We perform extensive CNN fine-tuning experiments and consolidate in one place the results for five different art-related classification tasks on three large fine art datasets. Along with addressing the previously explored tasks of artist, genre, style and time period classification, we introduce a novel task of classifying artworks based on their association with a specific national artistic context. We present state-of-the-art classification results of the addressed tasks, signifying the impact of our method on computational analysis of art, as well as other image classification related research areas. Furthermore, in order to question transferability of deep representations across various source and target domains, we systematically compare the effects of domain-specific weight initialization by evaluating networks pre-trained for different tasks, varying from object and scene recognition to sentiment and memorability labelling. We show that fine-tuning networks pre-trained for scene recognition and sentiment prediction yields better results than fine-tuning networks pre-trained for object recognition. This novel outcome of our work suggests that the semantic correlation between different domains could be inherent in the CNN weights. Additionally, we address the practical applicability of our results by analysing different aspects of image similarity. We show that features derived from fine-tuned networks can be employed to retrieve images similar in either style or content, which can be used to enhance capabilities of search systems in different online art collections.

With the emergence of large digitized fine art collections and the successful performance of deep learning techniques, new research prospects unfold in the intersection of artificial intelligence and art. In order to explore the applicability of deep learning techniques in understanding art images beyond object recognition and classification, we employ convolutional neural networks (CNN) to predict scores related to three subjective aspects of human perception: aesthetic evaluation of the image, sentiment evoked by the image, and memorability of the image. For each concept, we evaluate several different CNN models trained on various natural image datasets and select the best performing model based on the qualitative results and the comparison with existing subjective ratings of artworks. Furthermore, we employ different decision treebased machine learning models to analyze the relative importance of various image features related to the content, composition, and color in determining image aesthetics, visual sentiment, and memorability scores. Our findings suggest that content and image lighting have significant influence on aesthetics, in which color vividness and harmony strongly influence sentiment prediction, while object emphasis has a high impact on memorability. In addition, we explore the predicted aesthetic, sentiment, and memorability scores in the context of art history by analyzing their distribution in regard to different artistic styles, genres, artists, and centuries. The presented approach enables new ways of exploring fine art collections based on highly subjective aspects of art, as well as represents one step forward toward bridging the gap between traditional formal analysis and the computational analysis of fine art. INDEX TERMS Convolutional neural networks, image aesthetics, image memorability, fine art, visual sentiment.

No abstract

Real-world attacks can be interpreted as the result of competitive interactions between networks, ranging from predator -prey networks to networks of countries under economic sanctions. Although the purpose of an attack is to damage a target network, it also curtails the ability of the attacker, which must choose the duration and magnitude of an attack to avoid negative impacts on its own functioning. Nevertheless, despite the large number of studies on interconnected networks, the consequences of initiating an attack have never been studied. Here, we address this issue by introducing a model of network competition where a resilient network is willing to partially weaken its own resilience in order to more severely damage a less resilient competitor. The attacking network can take over the competitor's nodes after their long inactivity. However, owing to a feedback mechanism the takeovers weaken the resilience of the attacking network. We define a conservation law that relates the feedback mechanism to the resilience dynamics for two competing networks. Within this formalism, we determine the cost and optimal duration of an attack, allowing a network to evaluate the risk of initiating hostilities.