Computer vision and image recognition in archaeology
Claudia Engel, Peter Mangiafico, Justine Issavi, Dominik Lukas · 2019
We apply existing image recognition services (Google Vision API and Clarifai Predict API) for an image repository of about 150,000 archaeological images from the Çatalhöyük Archaeological Project. We train a custom machine learning model from labeled images with Google AutoML. We report on our initial explorations and discuss their usefulness for extracting and enriching metadata for discovery and research in the archaeological domain.