Archaeologists at Northern Arizona University used convolutional neural networks to teach computers to perform image-based classification of thousands of ancient pottery fragments into multiple stylistic categories rapidly and consistently.
"Using digital photographs of pottery, computers can accomplish what used to involve hundreds of hours of tedious, painstaking, and eye-straining work by archaeologists who physically sorted pieces of broken pottery into groups, in a fraction of the time and with greater consistency," says Leszek Pawlowicz of the Department of Anthropology. The work is described in "Applications of Deep Learning to Decorated Ceramic Typology and Classification: A Case Study Using Tusayan White Ware from Northeast Arizona," published in the Journal of Archaeological Science.
The researchers hope this approach to archaeological analysis of pottery can be applied to other types of ancient artifacts, moving archaeology into a phase of machine classification that results in greater efficiency of archaeological efforts.
From Northern Arizona University
View Full Article