This is a repository containing most of the documents sources I have redacted during my final year internship at the department of Archaeology of the University of Leiden, Netherlands. They are written in LaTeX. The base idea in this project is to use SoTA techniques in Satellite Imagery Detection on Archaeological LiDAR surveys detection. This proved very successful ! We will soon put out the code behind this framework, in an effort toward a more open science. Sadly, we cannot share the data, for multiple reasons.
Proposal for a research project on automated archaeological object detection in LiDAR surveys. Contains generalities about the projects, along with a planning, deliverables, and key dates.
A general State of The Art document recapitulating the SoTA on automated detection in Satellite Imagery, along with a shorter SoTA Archaeological Target Detection. It also contains a detailled look into YOLOv4.
A document proposing an Architecture for the detection framework. This architecture would be experimented with, but is generally based on YOLOv4
My end of study thesis. This documents presents some of the concepts and techniques used, along with the challenges of automated detection in Archaeology. A shorter, more concise SoTA is also presented. Finally, a deep look into the conception, experimentation and results is given. Statistics are computed and the framework is compared to the state of the art. Finally, some conclusions are given. If you just want to read one document here, read this one.
A presentation in Beamer for my final year defence. An even more concise introduction to everything I have done for 6 month. If you just want to get the basic gist of it, glean over this