-
3D Scene Understanding with GeoLLMs
Exploring approaches to build 3D-aware language models for geospatial scene understanding from multi-view imagery and 3D data.
-
Building a Satellite Image-Text Foundation Model with GeoCLIP
Adapting OpenCLIP for satellite imagery using large-scale datasets like LAION-400M and SatlasPretrain.
-
Property Intelligence from Satellite and Drone Imagery
Automated property analysis using segmentation models on satellite and drone imagery at varying resolutions.
-
Benchmarking Change Detection: Lessons from QFabric
Experimental results from benchmarking change detection, change type classification, and status tracking on the QFabric dataset.
-
Building a Multi-Task Change Detection Dataset
How we built QFabric, a multi-temporal satellite dataset for tracking urban construction changes across 100 cities.
