Transferware.AI: Automating the Identification of Transfer-Printed Ceramics Using Artificial Intelligence

Summary

This is an abstract from the session entitled "Paper / Report Submission (General Sessions)", at the 2025 annual meeting of the Society for Historical Archaeology.

When identifiable, transfer-printed ceramics provide detailed chronological, trade, and other cultural information for historic sites. However, there are thousands of distinct patterns from hundreds of manufacturers, and archaeologically recovered examples are often so fragmentary as to limit the ability to identify them. Yet, with several provisos this paper considers, those with the same pattern are identical and should be identifiable even from fragments. This paper reports on the progress of a new application of Artificial Intelligence and Machine Learning to this task. In partnership with the Transferware Collector’s Club and its database of more than 18,000 patterns, our team has developed a model in which uploaded pattern images output a matching category label from database. This model is implemented as an Artificial Neural Network, building a vector index out of the feature vectors of all images in the database and then extracting feature vectors for a submitted image for comparison.

This Resource is Part of the Following Collections

• Paper / Report Submission (General Sessions)   •
• Society for Historical Archaeology 2025

Keywords

Individual & Institutional Roles

Contact(s): Nicole Haddow