Digital Connoisseurship: Applications of Machine Learning to Moche Iconography

Summary

This is an abstract from the "SAA 2023: Individual Abstracts" session, at the 88th annual meeting of the Society for American Archaeology.

In the absence of a written language, the study of the complex narrative iconography of the Moche or Mochica culture of the North Coast of Perú (250-900CE) forms an important foundation of our understanding of the cultural dynamics and ritual traditions of this Pre-Columbian society. Fineline iconography on Moche ceramic vessels in museum and private collections in Perú and around the world continue to be the focus of considerable scholarship. However, in most archaeological contexts, only fragments of these vessels are found, allowing for an extremely partial understanding of the complete iconography of the original artifact. The process of assigning a fragment to a particular iconographic scene is entirely dependent on a researcher’s exposure to the artistic corpus, their “connoisseurship” or expert knowledge. The current project has developed machine learning techniques to assist in the recognition of iconographic scenes from fragmentary elements. An initial challenge involved creating an algorithm that “breaks” or “shatters” reference images into fragments for use in training a self-supervised deep learning system. Following training, an image of a novel fragment can be introduced to the system and assigned to a particular complete iconographic scenes with various degrees of confidence.

This Resource is Part of the Following Collections

• SAA 2023: Individual Abstracts   •
• Society for American Archaeology 88th Annual Meeting, Portland, OR (2023)

Keywords

Spatial Coverage

Record Identifiers

Abstract Id(s): 35934.0