---
license: mit
---

The pretrained models provided in this repo are trained from scratch using the [GeoAI](https://github.com/opengeos/geoai) Python package. 

- Building footprints detection for the USA using RGB images ([model checkpoint](https://huggingface.co/giswqs/geoai/blob/main/building_footprints_usa.pth) | [notebook](https://geoai.gishub.org/examples/train_building_footprints_usa))
- Building footprints detection for the USA using RGBN images ([model checkpoint](https://huggingface.co/giswqs/geoai/blob/main/building_footprints_usa_rgbn.pth) | [notebook](https://geoai.gishub.org/examples/train_object_detection_model))
- Wetland detection for the USA using RGBN images ([model checkpoint](https://huggingface.co/giswqs/geoai/blob/main/wetland_detection.pth) | [notebook](https://geoai.gishub.org/examples/wetland_mapping))
- Solar panel detection for the USA using RGB images ([model checkpoint](https://huggingface.co/giswqs/geoai/blob/main/solar_panel_detection.pth) | [notebook](https://geoai.gishub.org/examples/train_solar_panel_detection))
- Car detection for the USA using RGB images ([model checkpoint](https://huggingface.co/giswqs/geoai/blob/main/car_detection_usa.pth) | [notebook](https://geoai.gishub.org/examples/train_car_detection))
- Ship detection for the USA using RGB images ([model checkpoint](https://huggingface.co/giswqs/geoai/blob/main/ship_detection.pth) | [notebook](https://geoai.gishub.org/examples/train_ship_detection))
- Surface water mapping using RGBN images ([model checkpoint](https://huggingface.co/giswqs/geoai/blob/main/water_detection.pth) | [notebook](https://geoai.gishub.org/examples/train_water_detection))