Computer Assisted Ancient History

A lunchtime talk on the CDLI and DIEGEST projects applying imaging, machine learning and computer vision to cuneiform tablets, cylinder seals and ancient coin dies.

This lunchtime session presents two digital humanities projects. The Cuneiform Digital Library Initiative (CDLI) shares a pilot study on Old Babylonian cylinder seals, where expert annotation and an automatic detector trained on 2,491 examples identify human figures with 94% accuracy, and outlines the proposed CANVAS ERC project on seals, early scripts and symbol systems. DIEGEST then discusses automating die studies of ancient coins, including the migration of the tool from Singapore to a non-proprietary online environment at the Ashmolean Museum. The talk runs 13:15–14:00 in the Buttery, with tea, coffee and cakes provided, and an AWRC lunch table in Hall from 12:30 beforehand.