Information and code about the IGARSS 2022 publication
"TOWARDS DIGITAL SURFACE MODELS FOR DEEP LEARNING BUILDING CHANGE DETECTION".
The datasaets are kindly hosted by zenodo.org.
- 0.5m Ground Sampling Distance
- 0.2m Ground Sampling Distance
Use the script train.py to train any of the DDSM-based models presented in the paper.
Use the script test.py to test a trained and saved model.
If you find this work or the data itself useful, please cite us. ToDo: Add BiBTeX
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
