I am thrilled to announce the creation of a new dataset entitled ISSLIDE: InSAR dataset for Slow SLIding area DEtection available on IEEE Dataport.


The dataset is two-folds:
  • A raw interferogram dataset
  • A deep learning ready-to-be-used dataset
The raw dataset contains the whole interferograms as well as the shapefiles of the manually annotated landslides. Its objective is to be used for enrichment purpose or for the user to compute its own landslide patch extraction.
The ready-to-be-used dataset is already processed. 13.230 patches of phase difference and coherence were croped around the existing moves. The shapefiles are transformed into binary masks to facilitate deep learning applications.


Further details are available on IEEE Dataport and if you have any troubles with the dataset, do not hesitate to contact me.


If this dataset may be of any use for you, please consider citing it as follows:

Antoine Bralet, Emmanuel Trouvé, Jocelyn Chanussot, Abdourrahmane M. Atto, September 22, 2023, “ISSLIDE: InSAR dataset for Slow SLIding area DEtection with machine learning”, IEEE Dataport, doi: https://dx.doi.org/10.21227/dhxt-5g91.