Researchers at the Czech Technical University in the Czech Republic have developed a machine vision system that can analyze the poses of human subjects in fine art paintings throughout history, then identify other paintings that contain people in similar poses.
The method has revealed previously unknown links between art and artists, adding a powerful new tool that art historians can use to understand their field.
The system is based on the vast databases that include digitized collections from many of the world's top museums and galleries.
Said the researchers, "We experimentally show that explicit human pose matching is superior to standard content-based image retrieval methods on a manually annotated art composition transfer dataset."
From Technology Review
View Full Article
Abstracts Copyright © 2019 SmithBucklin, Washington, DC, USA