Project SIREN: Machine learning and the ancient naval battle site at the Egadi Islands, Sicily

Author(s): Mateusz Polakowski

Year: 2023

Summary

This is an abstract from the session entitled "Re-Visualizing Submerged Landscapes", at the 2023 annual meeting of the Society for Historical Archaeology.

On March 13, 241 BC off the western coast of Sicily a Roman naval force intercepted a Carthaginian resupply mission on its way to Sicily. Project SIREN is an endeavor to capture and use the experience gained through years of survey work on the Battle of the Egadi Islands Survey Project (2005-2021). Utilizing remote datasets including multibeam bathymetry (MBES), sidescan sonar (SSS), and video footage the project aims to apply machine learning in image recognition to the current survey methodology to increase the potential of identifying more warship remains. Early results indicate that image annotation can dramatically improve how we determine potential areas of investigation. This paper will discuss previous methods of investigation at the site, the experience gained in target recognition through direct survey methods, and how incorporation of automatic image identification is being used to increase the chances of locating artefacts related to the battle on the seabed.

Cite this Record

Project SIREN: Machine learning and the ancient naval battle site at the Egadi Islands, Sicily. Mateusz Polakowski. Presented at Society for Historical Archaeology, Lisbon, Portugal. 2023 ( tDAR id: 475823)

This Resource is Part of the Following Collections

Keywords

Geographic Keywords
Mediterranean

Individual & Institutional Roles

Contact(s): Nicole Haddow