Inferring compositional style in the neo-plastic paintings of Piet Mondrian by machine learning

Andrzejewski, David; Stork, David G.; Zhu, Xiaojin; Spronk, Ron

doi:10.1117/12.840558

Cited by 5 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each cell has a colour and three booleans that regulate growth: if hor is true, the cell is allowed to grow horizontally, if vert is true, it is allowed to grow vertically. If stops is true, it will stop when bumping into another cell, otherwise it will cross-over or cross through (in the terminology of Andrzejewski et al [1] a line that stops is a non-spanning line). Next, ratio is used to force the cell to have a certain height/width ratio but the special value −1 means that no ratio constraint applies.…”

Section: Working With Cellsmentioning

confidence: 99%

See 1 more Smart Citation

A program for Victory Boogie Woogie

Feijs

2019

Journal of Mathematics and the Arts

View full text Add to dashboard Cite

The paintings of Piet Mondrian have been attractive for programmers and researchers writing programs that produce compositions resembling the originals. In earlier work published in Leonardo, a generic framework was proposed, which turned out flexible and could be adapted to generate many different Mondrian types. Until recently, there were hardly any attempts to target Mondrian's Victory Boogie Woogie, which is more difficult than most of the earlier Mondrian types. This article describes an extension of the Leonardo approach by including Victory Boogie Woogie. This work poses different challenges because of its uniqueness, its complexity and the fact that it is unfinished. In the extension, three principles for modelling and programming are used: (1) working with cells, (2) nesting, and (3) object-orientation. The program entered and won a Dutch national competition on programming Victory Boogie Woogie in 2013.

show abstract

Section: Working With Cellsmentioning

confidence: 99%

“…Similarly, Andrzejewski et al [1] use feature vectors to train classifiers and generators. They use real Mondrians as training data, mostly the 1930-1940 compositions.…”

Section: Related Workmentioning

confidence: 99%

A program for Victory Boogie Woogie

Feijs

2019

Journal of Mathematics and the Arts

View full text Add to dashboard Cite

show abstract

“…Many paintings in the Western canon, particularly realist easel paintings from the Renaissance to the present, bear underdrawings and pentimenti-preliminary versions of the work created as the artist altered and developed into the final design. [1][2][3] In some cases the underdrawing represents a design separate from the final, visible work. Such ghost-paintings appear in the oeuvre of artists such as Pablo Picasso, Vincent van Gogh, Rembrandt, and Francisco Goya, among many others.…”

Section: Introductionmentioning

confidence: 99%

Resolution enhancement in the recovery of underdrawings via style transfer by generative adversarial deep neural networks

Cann¹,

Bourached²,

Griffiths³

et al. 2021

Preprint

View full text Add to dashboard Cite

We apply generative adversarial convolutional neural networks to the problem of style transfer to underdrawings and ghost-images in x-rays of fine art paintings with a special focus on enhancing their spatial resolution. We build upon a neural architecture developed for the related problem of synthesizing high-resolution photo-realistic image from semantic label maps. Our neural architecture achieves high resolution through a hierarchy of generators and discriminator sub-networks, working throughout a range of spatial resolutions. This coarse-to-fine generator architecture can increase the effective resolution by a factor of eight in each spatial direction, or an overall increase in number of pixels by a factor of 64. We also show that even just a few examples of human-generated image segmentations can greatly improve-qualitatively and quantitatively-the generated images. We demonstrate our method on works such as Leonardo's Madonna of the carnation and the underdrawing in his Virgin of the rocks, which pose several special problems in style transfer, including the paucity of representative works from which to learn and transfer style information.

show abstract

“…Computational geometric visualizations using the kd-tree data structure are involved in researches on the representation of Mondrian tables (LeFevre et al, 2006;Roy & Teh, 2008, Wang et al, 2015 on a three-dimensional cube (Skrodzki & Polthier, 2018). Andrzejewski et al (2010) focused on creating an educated productive model by using machine learning method based on the principles behind the compositions such as the intensity of use of colors to be able to recognize the original Mondrian artworks.…”

mentioning

confidence: 99%

Pixel-Based Geometric Decoding of Mondrian Compositions

Özgün¹,

Alaçam²

2021

Blucher Design Proceedings

View full text Add to dashboard Cite

This study explores the use of a computational method for decoding and encoding an art composition in the digital environment. Geometric decoding includes subdivision through color and shape. Five artworks of Mondrian, which contain simple but intense geometric data and have a unique composition style, are discussed in this context. Matrix-based expression of composition in terms of color and form with a code-based analysis allowed to definition of a control mechanism for existing geometric data. The aim of this decoding process, which follows segmentation and fragmentation methods, is to capture the common composition approach and particular comparison approach. Visual encoding, as the reverse of the decoding process, is the interpretation of the data obtained as a result of the analysis by assigning a color map. By this meanings, the matrixes and visual outputs enable the artworks to be compared according to the main composition style, or the similarities between them.

show abstract

Inferring compositional style in the neo-plastic paintings of Piet Mondrian by machine learning

Cited by 5 publications

References 8 publications

A program for Victory Boogie Woogie

A program for Victory Boogie Woogie

Resolution enhancement in the recovery of underdrawings via style transfer by generative adversarial deep neural networks

Pixel-Based Geometric Decoding of Mondrian Compositions

Contact Info

Product

Resources

About