Adapting Prithvi-EO for Fallow Detection for Food-Water Nexus: ViT-Adapter Necks and Parameter-Efficient Backbone tuning of Geospatial Foundation Model
arXiv:2606.12218v1 Announce Type: cross Abstract: Understanding spatial distribution of fallow land is important for optimizing the food-water (FW) nexus, given fallowing's role in crop rotation and water conservation. Fallow is a low accuracy class in USDA Cropland Data Layer (CDL). Geospatial foundation model (GFM), Prithvi-EO has shown strong transferability across computer vision tasks. However, its Vision Transformer (ViT) backbone produces features at a single spatial scale that are ill-suited for the multi-scale features required by object detection heads. Existing approaches synthesise
The increasing availability of geospatial foundation models and the urgent need for enhanced food security and water management are driving innovations in agricultural monitoring.
This development improves the accuracy of fallow land detection, which is critical for optimizing agricultural practices and managing water resources, directly impacting food-water nexus stability.
The ability to accurately identify fallow land using AI and satellite imagery allows for more efficient crop rotation planning and better water conservation strategies, moving beyond traditional, less precise methods.
- · Agriculture sector
- · Water management organizations
- · Geospatial AI companies
- · Governments
- · Traditional agricultural survey methods
- · Regions with inefficient water usage
Improved resource allocation for agricultural planning and water conservation.
Enhanced food security and reduced agricultural waste through data-driven decisions.
Potential for global agricultural optimization and resilience against climate change impacts on food and water availability.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI