Machine learning-ready remote sensing data for Maya archaeology: masks, ALS data, Sentinel-1, Sentinel-2
The dataset includes multimodal annotated data for remote sensing of Maya archaeology and is suitable for deep learning. The dataset covers the area around Chactún, one of the largest ancient Maya urban centres in the central Yucatán peninsula.
It includes five types of data:
- high-resolution airborne laser scanning (ALS, lidar) data visualisations (sky view factor, positive openness, slope),
- high-resolution airborne laser scanning derived canopy height model,
- Sentinel-1 Short Aperture Radar (SAR) satellite data (yearly average Sigma0),
- Sentine-2 optical satellite data (12 bands + cloud mask, 17 dates), and
- manual data annotations.
The manual annotations (used as binary masks) represent three different types of ancient Maya structures (class labels: buildings, platforms, and aguadas – artificial reservoirs) within the study area, their exact locations, and boundaries.
The dataset is ready for use with convolutional neural networks (CNNs) for object recognition, object localization (detection), and semantic segmentation. The dataset has already been used for the Discover the Mysteries of the Maya computer vision competition.
We would like to provide this dataset to help more research teams develop their own computer vision models for investigations of Maya archaeology or improve existing ones.
A detailed description of the datasets has been published by Kokalj, Ž., Džeroski, S., Šprajc, I. et al. Machine learning-ready remote sensing data for Maya archaeology. Scientific Data 10, 558 (2023). https://doi.org/10.1038/s41597-023-02455-x
The authors and institutions they are affiliated with exclude all liability for any reliance on the data.
