"An application of harmonic analysis to the digital humanities"Heinecke, AndreasDie analysis is an important tool of ancient economic history. Yet, manual die studies are too labor-intensive to comprehensively study large coinages such as those of the Roman Empire. From a computer vision viewpoint, such studies present a challenging unsupervised clustering problem, involving an unknown and large number of highly similar semantic classes (usually hundreds) of imbalanced sizes, for which training examples would be extremely time-expensive to obtain and are thus not available. We address this problem through determining dissimilarities derived from specifically devised re-weighted Gaussian process-based keypoint features in a Bayesian distance microclustering clustering framework. The covariance structure of the re-weighted Gaussian process can be efficiently approximated via convolutions. The resulting method can reduce the time investment necessary for large-scale die studies by several orders of magnitude, in many cases from years to weeks, and provide data towards longstanding controversial debates on the economic history of ancient states. |
« back