ADAF machine learning model weights

This dataset contains pretrained deep learning model weights for the ADAF (Automatic Detection of Archaeological Features) tool. Models are provided in PyTorch format and are compatible with the AiTLAS Python framework. They support both object detection (Faster R-CNN) and semantic segmentation (HRNet) tasks across four archaeological classes: all archaeology (AO), barrows, enclosures, and ringforts. These models were trained on expert-annotated airborne laser scanning (ALS) data from Ireland and are intended for use with the ADAF application. For installation and usage instructions, refer to the ADAF GitHub repository: https://github.com/EarthObservation/adaf.