Detecting visual design principles in art and architecture through deep convolutional neural networks

Gözdenur Demir; Aslı Çekmiş; Vahit Buğra Yeşilkaynak; Gozde Unal

doi:10.1016/j.autcon.2021.103826

Detecting visual design principles in art and architecture through deep convolutional neural networks

Gözdenur Demir, Aslı Çekmiş^*, Vahit Buğra Yeşilkaynak, Gozde Unal

^*Bu çalışma için yazışmadan sorumlu yazar

Yapay Zeka ve Veri Mühendisliği Bölümü

Istanbul Technical University

Araştırma sonucu: ???type-name??? › Makale › bilirkişi

18 Atıf (Scopus)

Özet

Visual design is associated with the use of some basic design elements and principles. Those are applied by the designers in the various disciplines for aesthetic purposes, relying on an intuitive and subjective process. Thus, numerical analysis of design visuals and disclosure of the aesthetic value embedded in them are considered as hard. However, it has become possible with emerging artificial intelligence technologies. This research aims at a neural network model, which recognizes and classifies the design principles over different domains. The domains include artwork produced since the late 20th century; professional photos; and facade pictures of contemporary buildings. The data collection and curation processes, including the production of computationally-based synthetic dataset, is genuine. The proposed model learns from the knowledge of myriads of original designs, by capturing the underlying shared patterns. It is expected to consolidate design processes by providing an aesthetic evaluation of the visual compositions with objectivity.

Orijinal dil	İngilizce
Makale numarası	103826
Dergi	Automation in Construction
Hacim	130
DOI'lar	https://doi.org/10.1016/j.autcon.2021.103826
Yayın durumu	Yayınlandı - Eki 2021

Bibliyografik not

Publisher Copyright:
© 2021 Elsevier B.V.

Finansman

We are grateful to the Museum of Contemporary Art Chicago for providing copyrights of the artworks, and following photographers and architects for granting permission to use their photographs: Aitor Ortiz, Tord-Rikard Söderström, Katja L.(Instagram/@architecturetourist), Altug Galip (Instagram/@kyrenian), Florent Michel, René de Wit, Tomás Alvarez Robledo, John Gollings, Ieva Saudargaité, Miguel Souto, SOMOS.Arquitectos, and Peter Kuczia. We are thankful to Cemil Cahit Yavuz, the artist and his team for data labeling. We wish to thank Murat Can Kurşun, the graphic designer; Prof.Dr. Yüksel Demir, the architect and; Assoc.Prof.Dr. Oğuz Haşlakoğlu, the artist who put their expert knowledge and annotated the new dataset in the human vs. computer competition experiment. We also wish to thank Gülçin Baykal, Dilara Gökçe and Ahmed Hancıoğlu for their great contributions in data curation. This project was financially supported by the Scientific Research Project Unit (BAP) of Istanbul Technical University, Project Number: MOA-2019-42321. We are grateful to the Museum of Contemporary Art Chicago for providing copyrights of the artworks, and following photographers and architects for granting permission to use their photographs: Aitor Ortiz, Tord-Rikard Söderström, Katja L.(Instagram/@architecturetourist), Altug Galip (Instagram/@kyrenian), Florent Michel, René de Wit, Tomás Alvarez Robledo, John Gollings, Ieva Saudargaité, Miguel Souto, SOMOS.Arquitectos, and Peter Kuczia. We are thankful to Cemil Cahit Yavuz, the artist and his team for data labeling. We wish to thank Murat Can Kurşun, the graphic designer; Prof.Dr. Yüksel Demir, the architect and; Assoc.Prof.Dr. Oğuz Haşlakoğlu, the artist who put their expert knowledge and annotated the new dataset in the human vs. computer competition experiment. We also wish to thank Gülçin Baykal, Dilara Gökçe and Ahmed Hancıoğlu for their great contributions in data curation. This project was financially supported by the Scientific Research Project Unit (BAP) of Istanbul Technical University , Project Number: MOA-2019-42321 .

Finansörler	Finansör numarası
Altug Galip
Cemil Cahit Yavuz
Murat Can Kurşun
Istanbul Teknik Üniversitesi	MOA-2019-42321

Belgeye Erişim

10.1016/j.autcon.2021.103826

Diğer Dosyalar ve Linkler

Link to publication in Scopus

Alıntı Yap

@article{ad110c28e2c74bdcbb6ad5c837e737bf,

title = "Detecting visual design principles in art and architecture through deep convolutional neural networks",

abstract = "Visual design is associated with the use of some basic design elements and principles. Those are applied by the designers in the various disciplines for aesthetic purposes, relying on an intuitive and subjective process. Thus, numerical analysis of design visuals and disclosure of the aesthetic value embedded in them are considered as hard. However, it has become possible with emerging artificial intelligence technologies. This research aims at a neural network model, which recognizes and classifies the design principles over different domains. The domains include artwork produced since the late 20th century; professional photos; and facade pictures of contemporary buildings. The data collection and curation processes, including the production of computationally-based synthetic dataset, is genuine. The proposed model learns from the knowledge of myriads of original designs, by capturing the underlying shared patterns. It is expected to consolidate design processes by providing an aesthetic evaluation of the visual compositions with objectivity.",

keywords = "Computer vision, Deep convolutional neural network (CNN), Deep learning, Image recognition, Visual analysis, Visual design principles",

author = "G{\"o}zdenur Demir and Aslı {\c C}ekmi{\c s} and Ye{\c s}ilkaynak, {Vahit Buğra} and Gozde Unal",

note = "Publisher Copyright: {\textcopyright} 2021 Elsevier B.V.",

year = "2021",

month = oct,

doi = "10.1016/j.autcon.2021.103826",

language = "English",

volume = "130",

journal = "Automation in Construction",

issn = "0926-5805",

publisher = "Elsevier B.V.",

}

TY - JOUR

T1 - Detecting visual design principles in art and architecture through deep convolutional neural networks

AU - Demir, Gözdenur

AU - Çekmiş, Aslı

AU - Yeşilkaynak, Vahit Buğra

AU - Unal, Gozde

PY - 2021/10

Y1 - 2021/10

N2 - Visual design is associated with the use of some basic design elements and principles. Those are applied by the designers in the various disciplines for aesthetic purposes, relying on an intuitive and subjective process. Thus, numerical analysis of design visuals and disclosure of the aesthetic value embedded in them are considered as hard. However, it has become possible with emerging artificial intelligence technologies. This research aims at a neural network model, which recognizes and classifies the design principles over different domains. The domains include artwork produced since the late 20th century; professional photos; and facade pictures of contemporary buildings. The data collection and curation processes, including the production of computationally-based synthetic dataset, is genuine. The proposed model learns from the knowledge of myriads of original designs, by capturing the underlying shared patterns. It is expected to consolidate design processes by providing an aesthetic evaluation of the visual compositions with objectivity.

AB - Visual design is associated with the use of some basic design elements and principles. Those are applied by the designers in the various disciplines for aesthetic purposes, relying on an intuitive and subjective process. Thus, numerical analysis of design visuals and disclosure of the aesthetic value embedded in them are considered as hard. However, it has become possible with emerging artificial intelligence technologies. This research aims at a neural network model, which recognizes and classifies the design principles over different domains. The domains include artwork produced since the late 20th century; professional photos; and facade pictures of contemporary buildings. The data collection and curation processes, including the production of computationally-based synthetic dataset, is genuine. The proposed model learns from the knowledge of myriads of original designs, by capturing the underlying shared patterns. It is expected to consolidate design processes by providing an aesthetic evaluation of the visual compositions with objectivity.

KW - Computer vision

KW - Deep convolutional neural network (CNN)

KW - Deep learning

KW - Image recognition

KW - Visual analysis

KW - Visual design principles

UR - http://www.scopus.com/inward/record.url?scp=85111539313&partnerID=8YFLogxK

U2 - 10.1016/j.autcon.2021.103826

DO - 10.1016/j.autcon.2021.103826

M3 - Article

AN - SCOPUS:85111539313

SN - 0926-5805

VL - 130

JO - Automation in Construction

JF - Automation in Construction

M1 - 103826

ER -

Detecting visual design principles in art and architecture through deep convolutional neural networks

Özet

Bibliyografik not

Finansman

Belgeye Erişim

Diğer Dosyalar ve Linkler

Parmak izi

Alıntı Yap