Search

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. [ABSTRACT FROM AUTHOR]

Selenium (Se) is a trace element that can cause human physical deformations due to selenosis, and mutagenic changes on a range of sensitive macro-organisms. Previous studies indicated that Se enrichment is significantly greater in coal than in other rock units, suggesting a correlation between Se and coal bed proximity. Since coal extraction can cause a release of Se, some regulatory authorities have responded by requiring sampling of coal seams and adjacent rock for Se prior to issuing coal mining permits. This sampling is done under the assumption that Se in a small number of samples will be correlated to the Se concentrations throughout the strata they represent. For example, in one case a single continuous deep rock core was sampled for Se to estimate the concentrations of Se across a 2.52 km2 proposed surface coal mine. This project examined the adequacy of such an approach using univariate statistics and variograms to determine the spatial variability and correlation of Se concentrations collected from six Appalachian coal seams. The results from all analyses showed no significant spatial correlation of Se within any coal seam examined. Given this lack of measured Se spatial correlation in Appalachian coal seams for scales as small as 0.10 km, it is unlikely that a Se spatial correlation exists in adjacent rock units at similar scales either, which is currently the assumption being made for sampling to minimize Se pollution from surface coal mining. [ABSTRACT FROM AUTHOR]