3 models
Detection
Find and locate individual objects in imagery, from buildings and vehicles to ships and trees.
Geospatial AI, explained
Explore GeoAI, one model category at a time
Understand the task, explore model families and learn how to train and use them. Start with a category, then discover practical guides, image examples and model resources.
13 task categories · 26 published models
Choose your task
GeoAI model categories
Read the category guide or browse its available models.
3 models
Classification
Assign class labels to image tiles or feature samples, such as water, vegetation or built-up land.
3 models
Semantic Segmentation
Create a class map by assigning a label to every pixel, such as water, road, vegetation or building.
2 models
Instance Segmentation
Trace each object with its own mask, keeping neighbouring buildings, trees or other instances separate.
3 models
Change Detection
Compare the same place across dates to find construction, vegetation loss or other meaningful changes.
3 models
Regression
Predict numerical quantities such as canopy height, biomass or another measured environmental variable.
4 models
Point Cloud Segmentation
Label 3D points in LiDAR and survey clouds to distinguish terrain, buildings, vegetation and other scene classes.
1 model
3D Object Detection
Locate objects in three-dimensional space using LiDAR or other depth measurements.
1 model
Time-Series Forecasting
Predict future sensor values and their uncertainty from time-stamped observations.