Remote Sensing

The field turns optical, radar, hyperspectral, thermal, and LiDAR observations into maps and measurements of land cover, buildings, forests, hazards, weather, and environmental change. Common approaches include convolutional and transformer models, time-series analysis, multimodal sensor fusion, geospatial foundation models, and methods that incorporate physical knowledge or quantify uncertainty. Researchers evaluate these systems on tasks such as segmentation, detection, image restoration, retrieval, and forecasting, increasingly using geographically separated tests and benchmarks that reflect real operating conditions. A central challenge is making results reliable across regions, sensors, seasons, and limited-label settings while balancing accuracy with interpretability, uncertainty, and onboard or large-scale computational constraints.