Computational Ceramicology

TL;DR

This paper introduces two machine learning tools for archeological pottery identification, utilizing shape and decoration analysis with novel architectures and training methods to address real-world data challenges.

Contribution

It presents innovative deep learning architectures and training strategies tailored for archeological data, improving pottery classification accuracy with limited and imbalanced datasets.

Findings

Effective shape-based classification using a novel deep-learning architecture.

Successful decoration feature recognition with standard image recognition models.

Overcame data scarcity and class imbalance through synthetic data and specialized loss functions.

Abstract

Field archeologists are called upon to identify potsherds, for which purpose they rely on their experience and on reference works. We have developed two complementary machine-learning tools to propose identifications based on images captured on site. One method relies on the shape of the fracture outline of a sherd; the other is based on decorative features. For the outline-identification tool, a novel deep-learning architecture was employed, one that integrates shape information from points along the inner and outer surfaces. The decoration classifier is based on relatively standard architectures used in image recognition. In both cases, training the classifiers required tackling challenges that arise when working with real-world archeological data: paucity of labeled data; extreme imbalance between instances of the different categories; and the need to avoid neglecting rare classes and to take note of minute distinguishing features of some classes. The scarcity of training data was overcome by using synthetically-produced virtual potsherds and by employing multiple data-augmentation techniques. A novel form of training loss allowed us to overcome the problems caused by under-populated classes and non-homogeneous distribution of discriminative features.