This code was used to produce the results shown in:
Mabon, J.; Ortner, M.; Zerubia, J. Learning Point Processes and Convolutional Neural Networks for Object Detection in Satellite Images. Remote Sens. 2024, 16, 1019. https://doi.org/10.3390/rs16061019
If you use this code, we strongly suggest you cite:
@Article{rs16061019,
AUTHOR = {Mabon, Jules and Ortner, Mathias and Zerubia, Josiane},
TITLE = {Learning Point Processes and Convolutional Neural Networks for Object Detection in Satellite Images},
JOURNAL = {Remote Sensing},
VOLUME = {16},
YEAR = {2024},
NUMBER = {6},
ARTICLE-NUMBER = {1019},
URL = {https://www.mdpi.com/2072-4292/16/6/1019},
ISSN = {2072-4292},
DOI = {10.3390/rs16061019}
}
For further details you can refer to:
Jules Mabon. Learning stochastic geometry models and convolutional neural networks. Application to multiple object detection in aerospatial data sets. Signal and Image Processing. Université Côte d'Azur, 2023. English. https://theses.hal.science/tel-04404849
- setup conda environment with
conda env create -f env.yml
- setup models and data paths with paths_config.yaml.
- to compute metrics install dota devkit in
data/
(see installation for more info)
cd data/
git clone https://github.com/CAPTAIN-WHU/DOTA_devkit
cd DOTA_devkit/
swig -c++ -python polyiou.i
python setup.py build_ext --inplace
- configure
paths_configs.json
as needed - conda env is provided
env.yml
, setup usingconda env create -f env.yml
Results with model mppe_dota_LMM_0305d on sample data from ADS.
Results details, with right plot showing to data and prior scores correspondence to color
- saved_models: saved models that can be used for inference (use setup in paths_config.yaml)
- env.yml: file defining the conda environment suitable to run the code
- model_configs: config files for models
# training (use -o to overwrite existing model)
python main.py -p train mppe_dota_LMM_305d.yaml
# inferrence and evalution
python main.py -p inferval mppe_dota_LMM_305d
# making some figures
python tools/figures_results.py
# compute subscores
python tools/compute_subscores.py
# show subscores resutls
python tools/figures_subscores.py
- Minitel font (CC0 license)