FireScope trains a VLM on US data to output wildfire risk rasters with reasoning traces and shows improved cross-continental performance on European events compared with prior approaches.
SMLFire1.0: a stochastic machine learning model for fire frequency and size distributions across the western united states.Geoscientific Model Development, 16:3407–3432, 2023
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
FireScope: Wildfire Risk Raster Prediction with a Chain-of-Thought Oracle
FireScope trains a VLM on US data to output wildfire risk rasters with reasoning traces and shows improved cross-continental performance on European events compared with prior approaches.