Machine and deep learning

Landscape Archaeology Research Group

We aim to develop workflows and open algorithms that can help and eventually improve detection, identification, mapping and quantification of features of archaeological interest. These include not only sites but also different types of archaeological remains.
Current methods include machine learning, deep learning and approaches such as pattern matching, network analysis, high-resolution 3D reconstructions and geometric morphometrics.

Deep learning-based automated detection of tumuli on a MSRM raster derived from filtered lidar data