This repo is an implementation code for a conference paper Variational Autoencoders for Unsupervised Object Counting from VHR Imagery: Applications in Dwelling Extraction from Forcibly Displaced People Settlement Areas. The experiment used multisource, multi-site very high-resolution satellite imagery obtained from different geographies hosting forcibly displaced people. The workflow is indicated in the following figure.

Usage

training sh ae_train.sh

Testing

sh ae_test.sh

Citation

if you use this paper, please cite as:

@inproceedings{gella2023variational,
  title={Variational Autoencoders for Unsupervised Object Counting from VHR Imagery: Applications in Dwelling Extraction from Forcibly Displaced People Settlement Areas},
  author={Gella, Getachew Workineh and Gangloff, Hugo and Wendt, Lorenz and Tiede, Dirk and Lang, Stefan},
  booktitle={IGARSS 2023-2023 IEEE International Geoscience and Remote Sensing Symposium},
  pages={1162--1165},
  year={2023},
  organization={IEEE}
}