Learning Algorithms for Digital Reconstruction of Van Gogh’s Drawings

Conference paper (2016)

Authors

J. Tang Pattern Recognition and Bioinformatics - EEMCS
J.C.A. van der Lubbe Cyber Security - EEMCS
M. Loog Pattern Recognition and Bioinformatics - EEMCS
Machine learning Van Gogh’s drawing Drawing reconstruction Reproduction
More Info
expand_more

Published Date

2016

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Many works of Van Gogh’s oeuvre, such as letters, drawings and paintings, have been severely degraded due to light exposure. Digital reconstruction of faded color can help to envisage how the artist’s work may have looked at the time of creation. In this paper, we study the reconstruction of Vincent van Gogh’s drawings by means of learning schemes and on the basis of the available reproductions of these drawings. In particular, we investigate the use of three machine learning algorithms, k-nearest neighbor (kNN) estimation, linear regression (LR), and convolutional neural networks (CNN), for learning the reconstruction of these faded drawings. Experimental results show that the reconstruction performance of the kNN method is slightly better than those of the CNN. The reconstruction performance of the LR is much worse than those of the kNN and the CNN.