---
license: cc0-1.0
pipeline_tag: image-segmentation
---

This repository contains three fine-tuned [TerraMind](https://github.com/IBM/terramind) models, using ```terramind_v1_base``` backbone. They are able to segment respectively seismic fault scarps, paleo-landslides, and river active channels, based on Sentinel-2 RGB and infrared images as well as a Digital Elevation Model (DEM). A UPerNet decoder is added to the TerraMind backbone, and the fine-tuning is done on three released datasets, one for each object (c.f. GitLab repository).

- Full repository: [GitLab](https://gricad-gitlab.univ-grenoble-alpes.fr/fauref/geomorphological-mapping-terramind).
- The seismic fault scarp fine-tuned model is avalaible here: [TerraMindUPerNet_sfs.pth](https://huggingface.co/francoisfaure/Geomorphological_Mapping_TerraMind/blob/main/TerraMindUPerNet_sfs.pth)
- The paleo-landslide fine-tuned model is avalaible here: [TerraMindUPerNet_pls.pth](https://huggingface.co/francoisfaure/Geomorphological_Mapping_TerraMind/blob/main/TerraMindUPerNet_pls.pth)
- The river active channel fine-tuned model is avalaible here: [TerraMindUPerNet_rac.pth](https://huggingface.co/francoisfaure/Geomorphological_Mapping_TerraMind/blob/main/TerraMindUPerNet_rac.pth)