Introduction

Read the abstract and introduction at the beginning here about the dataset: * https://science.sciencemag.org/content/361/6409/eaau0137

What is this project about?

• We are collaborating with team of archeologists to process a Guatemala dataset with DL methods
• The task is very complex, You will be assigned a small subset of the problems to test, implement and evaluate.
• Selecting the most suitable tile for segmentation.The dataset contains multiple tiles with size (approx 17100 x 9800 pixels), that represents different transforms of the lidar source data
• Expected output: Multiple models, that perform well on the selected dataset for detection and segmentation.

SOA – Image classification

Convolutional Neural Network Predictions
:

Tools and technologies

• Python (pytorch, numpy, open cv…), Matlab (computer vision and deep learning toolboxes)
• QGIS

Literature

About the methods used:

• https://lmb.informatik.uni-freiburg.de/people/ronneber/u-net/
• https://blog.athelas.com/a-brief-history-of-cnns-in-image-segmentation-from-r-cnn-to-mask-r-cnn-34ea83205de4
• https://pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html

Contact information

• miroslav.jascur@tuke.sk
• marek.bundzel@tuke.sk