Statistics, vision, and the analysis of artistic style

Graham, Daniel J.; Hughes, James M.; Leder, Helmut; Rockmore, Daniel N.

doi:10.1002/wics.197

Cited by 24 publications

(11 citation statements)

References 36 publications

(52 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Crowdsourcing has also been used to collect large databases of human perceptions of city images such as ‘safety’, ‘beauty’ and ‘happiness’ [ 20 , 21 ]. Computer vision methods such as ‘sparse coding’ [ 22 ] and ‘bag of visual words’ [ 23 ] have allowed researchers to identify statistical characteristics and specific areas of images that relate to concepts such as ‘artistic style’ [ 24 ] or visual perceptions of cities [ 25 ]. More recently, the introduction of convolutional neural networks (CNNs) has led to dramatic improvements in computer vision tasks, including visual recognition [ 26 , 27 ], understanding image aesthetics [ 28 , 29 ] and extracting perceptions of urban neighbourhoods [ 30 , 31 ].…”

Section: Introductionmentioning

confidence: 99%

Using deep learning to quantify the beauty of outdoor places

2017

View full text Add to dashboard Cite

Beautiful outdoor locations are protected by governments and have recently been shown to be associated with better health. But what makes an outdoor space beautiful? Does a beautiful outdoor location differ from an outdoor location that is simply natural? Here, we explore whether ratings of over 200 000 images of Great Britain from the online game Scenic-Or-Not, combined with hundreds of image features extracted using the Places Convolutional Neural Network, might help us understand what beautiful outdoor spaces are composed of. We discover that, as well as natural features such as ‘Coast’, ‘Mountain’ and ‘Canal Natural’, man-made structures such as ‘Tower’, ‘Castle’ and ‘Viaduct’ lead to places being considered more scenic. Importantly, while scenes containing ‘Trees’ tend to rate highly, places containing more bland natural green features such as ‘Grass’ and ‘Athletic Fields’ are considered less scenic. We also find that a neural network can be trained to automatically identify scenic places, and that this network highlights both natural and built locations. Our findings demonstrate how online data combined with neural networks can provide a deeper understanding of what environments we might find beautiful and offer quantitative insights for policymakers charged with design and protection of our built and natural environments.

show abstract

Section: Introductionmentioning

confidence: 99%

Using deep learning to quantify the beauty of outdoor places

2017

View full text Add to dashboard Cite

show abstract

“…Related research employs digital image analysis techniques to study the image features and image statistics that underwrite our capacity to recognize the subjects of pictorial representations and sort them into stylistic categories (Bonnar, Gosselin, and Schyns ; Greene and Oliva ; Graham et al. ). This research has been used both to develop digital image analysis techniques and to further our understanding of the nature of artistic style.…”

Section: The Codependence Of Scientific and Artistic Understanding Inmentioning

confidence: 99%

“…A central assumption of neuroaesthetics is that artworks, like behavioral deficits studied in neuropsychology, reflect, and so can function as tools for revealing, facts about the neurophysiological and psychological mechanisms underlying ordinary perception (Zeki 1999;Cavanagh 2005). Related research employs digital image analysis techniques to study the image features and image statistics that underwrite our capacity to recognize the subjects of pictorial representations and sort them into stylistic categories (Bonnar, Gosselin, and Schyns 2002;Greene and Oliva 2009;Graham et al 2012). This research has been used both to develop digital image analysis techniques and to further our understanding of the nature of artistic style.…”

Section: The Dependence Of Artistic Creativity On Scientific Innmentioning

confidence: 99%

Art and Science: A Philosophical Sketch of Their Historical Complexity and Codependence

Bullot

Seeley

Davies

2017

The Journal of Aesthetics and Art Criticism

View full text Add to dashboard Cite

To analyze the relations between art and science, philosophers and historians have developed different lines of inquiry. A first type of inquiry considers how artistic and scientific practices have interacted over human history. Another project aims to determine the contributions (if any) that scientific research can make to our understanding of art, including the contributions that cognitive science can make to philosophical questions about the nature of art. We rely on contributions made to these projects in order to demonstrate that art and science are codependent phenomena. Specifically, we explore the codependence of art and science in the context of a historical analysis of their interactions and in the context of contemporary debates on the cognitive science of art.

show abstract

“…It allowed them to investigate the richness of vocabulary exhibited in these texts under the proposition that the writing style usually varies depending on the targeted readership or audience. Graham et al (2012) state that in literature, there is an established set of techniques that have been successfully leveraged in the statistical analysis of literary style. The most common purpose of the analysis is to answer questions of authenticity and attribution.…”

Section: Introductionmentioning

confidence: 99%

Classification of the Educational Texts Styles With the Methods of Artificial Intelligence

Horáková

Houška

Dömeová

2017

JBSE

View full text Add to dashboard Cite

Modern educational methods emphasize the necessity to transfer knowledge instead of data or information within the educational process. Thus it is important to the educational texts supporting the educational process contain knowledge in a particular textual representation. But it is not trivial to decide whether the particular piece of text contain knowledge or not. The solution is to measure the similarity between the particular text structure and a typical structure of a knowledge-designed text. This research aims at analysing the classification ability of three commonly-used classification techniques: artificial neural networks (ANNs), classification and regression trees (CARTs) and decision trees (bigMLs) to separate texts or text fragments into two groups. The texts in the first group contain mainly data and information (common texts), the texts in the other group contain knowledge in one of the particular knowledge representations (knowledge texts). The sample of 120 text fragments was used for the analysis. The results show that the ANN techniques are significantly more able to make the right classification of the text than the CART or bigML ones, and evidence good classification abilities. Thus the ANN approach could broaden the set of methods used for evaluation of difficulty of educational texts or textbooks. Keywords: artificial intelligence, classification and regression trees, educational texts, knowledge representation, knowledge unit, production rules, stylometric analysis.

show abstract

Statistics, vision, and the analysis of artistic style

Cited by 24 publications

References 36 publications

Using deep learning to quantify the beauty of outdoor places

Using deep learning to quantify the beauty of outdoor places

Art and Science: A Philosophical Sketch of Their Historical Complexity and Codependence

Classification of the Educational Texts Styles With the Methods of Artificial Intelligence

Contact Info

Product

Resources

About