Authors:
Eric Piette, Achille Morenville, Barbara Care, Dorina Moullou, Walter Crist

Venue:
Computer Applications and Quantitative Methods in Archaeology (CAA), 2025

Topics:
game recognition, computer vision, supervised learning, dataset, Ludii, cultural heritage

Links: PDF

This work introduces a collaborative dataset of board game images designed to support AI-based recognition of traditional games.

The dataset aggregates photographs of game boards from diverse cultural and historical contexts, annotated by experts to enable supervised learning approaches. These data are used to train models capable of identifying and ranking candidate games from the Ludii game database.

The ultimate goal is to develop an application that allows users to upload images of board games and receive likely matches, supporting archaeologists, researchers, and the general public in identifying and understanding game artefacts.

This research represents a key step toward integrating computer vision and cultural heritage within the GameTable initiative.

By combining collaborative data collection with supervised learning techniques such as convolutional neural networks, the project enables automatic recognition of games from images—an important challenge in archaeology and heritage studies.

The work also extends the capabilities of the Ludii ecosystem, transforming it into a central hub not only for modelling and analysing games, but also for identifying and linking real-world artefacts to computational representations.

Piette, E., Morenville, A., Care, B., Moullou, D., Crist, W. (2025). AI-powered Game Recognition: A Collaborative Dataset for Traditional Games. Computer Applications and Quantitative Methods in Archaeology (CAA).

@inproceedings{piette2025game_recognition,
  author    = {Piette, Eric and Morenville, Achille and Care, Barbara and Moullou, Dorina and Crist, Walter},
  title     = {AI-powered Game Recognition: A Collaborative Dataset for Traditional Games},
  booktitle = {Computer Applications and Quantitative Methods in Archaeology (CAA)},
  year      = {2025}
}