Abstract
Attribution of paintings is a critical problem in art history. This study extends machine learning analysis to surface topography of painted works. A controlled study of positive attribution was designed with paintings produced by a class of art students. The paintings were scanned using a chromatic confocal optical profilometer to produce surface height data. The surface data were divided into virtual patches and used to train an ensemble of convolutional neural networks (CNNs) for attribution. Over a range of square patch sizes from 0.5 to 60 mm, the resulting attribution was found to be 60–96% accurate, and, when comparing regions of different color, was nearly twice as accurate as CNNs using color images of the paintings. Remarkably, short length scales, even as small as a bristle diameter, were the key to reliably distinguishing among artists. These results show promise for real-world attribution, particularly in the case of workshop practice.
Highlights
Machine learning (ML) analysis for artwork is a budding methodology aimed at advancing connoisseurship, the primary method of determining the attribution of an artwork, among other applications involving artistic style
A controlled study of paintings commissioned from several artists that mimic certain workshop practices are interrogated for indicative stylistic information. The aim of this experiment is to explore the questions that (a) the brushstroke-produced high resolution profilometry data from a painting’s surface contain stylistic information, and (b) the data are such that ML analysis can quantitatively distinguish among painters by the topographical information
Each patch is color coded according to the highest probability, with the opacity of the shading proportional to the magnitude of that probability
Summary
Machine learning (ML) analysis for artwork is a budding methodology aimed at advancing connoisseurship, the primary method of determining the attribution of an artwork, among other applications involving artistic style. A controlled study of paintings commissioned from several artists that mimic certain workshop practices are interrogated for indicative stylistic information The aim of this experiment is to explore the questions that (a) the brushstroke-produced high resolution profilometry data from a painting’s surface contain stylistic information (i.e., painters leave behind a measurable “fingerprint”), and (b) the data are such that ML analysis can quantitatively distinguish among painters by the topographical information. The experimental goal is to categorize small areas from the surface of paintings by their stylometric information, without the influence of purposeful stylistic choices (e.g., tools or materials) or factors regarding the subject of the painting To this effect, a series of twelve paintings by four artists, and their associated topographical profiles, are subject to analysis to attribute the works and to ascertain the important properties involved in those attributions. That the methods described here could find application elsewhere, such as forgery detection in contemporary art
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
