Adapting Prithvi-EO for Fallow Detection for Food-Water Nexus: ViT-Adapter Necks and Parameter-Efficient Backbone tuning of Geospatial Foundation Model
Researchers have developed a new method to improve the detection of fallow agricultural land, which is crucial for optimizing food and water resources. Their approach adapts the Prithvi-EO geospatial foundation model using parameter-efficient fine-tuning techniques like LoRA and ViT-Adapter necks. This method significantly enhances the model's ability to capture localized fallow patterns, outperforming existing techniques by up to 25.70% in detection accuracy. AI
IMPACT Enhances agricultural monitoring capabilities by improving the accuracy of fallow land detection, aiding in food-water nexus optimization.