Seeing Constable’s Clouds: An Application of Machine Learning to Art Historical Research

FAIN: HAA-271801-20

Pennsylvania State University (University Park, PA 16802-1503)
Elizabeth C. Mansfield (Project Director: January 2020 to April 2025)
James Z. Wang (Co Project Director: May 2020 to April 2025)

The development of computational methods to analyze formal details in paintings, focusing on cloud studies by John Constable and his emulators, documentary photographs, and fine art photographs.  

“Seeing Constable’s Clouds” proposes to use computer vision and machine learning to better understand the visual cues of 19th-century pictorial realism. Realist landscapes by British artist John Constable (1776-1837) are often split into two phases depending on whether they were made before or after an intense period spent observing and painting clouds in 1821-1822. Contemporaries disagreed about the effect this interlude had on the realism of his work, and art historians continue to debate how his empirical approach influenced his style and technique. The project team will use computer vision to seek formal details that art historians may have overlooked or been unable to discern. We will also use machine learning to discover whether there are formal features that contribute to the verisimilitude of three different types of pictorial realism: Constable’s paintings of clouds, fine art photographs of clouds, and research photographs of clouds made for scientific study.

Media Coverage

Can Machines Do Art History? (Media Coverage)
Author(s): Jo Lawson-Tancred
Publication: Apollo
Date: 12/3/2021
URL: https://www.apollo-magazine.com/ai-machine-learning-art-history/

Associated Products

Techniques of the Art Historical Observer (Article)
Title: Techniques of the Art Historical Observer
Author: Elizabeth Mansfield
Author: Zhuomin Zhang
Author: Jia Li
Author: John Russell
Author: George S. Young
Author: Catherine Adams
Author: James Z. Wang
Abstract: Computer vision has proven a useful tool for the technical analysis of art, yielding valuable information about condition, artistic processes, and materials, as well as insights into provenance and authenticity. When it comes to art historical interpretation—i.e., the diverse practices used to elicit historical meaning from works of visual or material culture—computer vision has been applied more tentatively, at least by art historians. This essay attempts to account for this hesitancy by providing an account of the Seeing Constable’s Clouds project from the perspective of one member of the research team, the art historian.
Year: 2022
Primary URL: https://www.19thc-artworldwide.org/spring22/practicing-art-history-techniques-of-the-art-historical-observer
Primary URL Description: Link to article in Nineteenth-Century Art Worldwide
Access Model: open access
Format: Journal
Periodical Title: Nineteenth-Century Art Worldwide
Publisher: Association of Historians of Nineteenth-Century Art

A Machine Learning Paradigm for Studying Pictorial Realism: Are Constable's Clouds More Real than His Contemporaries? (Database/Archive/Digital Edition)
Title: A Machine Learning Paradigm for Studying Pictorial Realism: Are Constable's Clouds More Real than His Contemporaries?
Author: Zhuomin Zhang
Author: Elizabeth Mansfield
Author: Jia Li
Author: John Russell
Author: George S. Young
Author: Catherine Adams
Author: James Z. Wang
Abstract: European artists have sought to create life-like images since the Renaissance. The techniques used by artists to impart realism to their paintings often rely on approaches based in mathematics, like linear perspective; yet the means used to assess the verisimilitude of realist paintings have remained subjective, even intuitive. An exploration of alternative and relatively objective methods for evaluating pictorial realism could enhance existing art historical research. We propose a machine-learning-based paradigm for studying pictorial realism in an explainable way. Unlike subjective evaluations made by art historians or computer-based painting analysis exploiting inexplicable learned features, our framework assesses realism by measuring the similarity between clouds painted by exceptionally skillful 19th-century landscape painters like John Constable and photographs of clouds. The experimental results of cloud classification show that Constable approximates more consistently than his contemporaries the formal features of actual clouds in his paintings. Our analyses suggest that artists working in the decades leading up to the invention of photography worked in a mode that anticipated some of the stylistic features of photography. The study is a springboard for deeper analyses of pictorial realism using computer vision and machine learning.
Year: 2022
Primary URL: https://arxiv.org/abs/2202.09348
Primary URL Description: Link to arXiv where draft article has been deposited.
Access Model: open access

---