AI-powered identification of tropical tree seedlings using hyperspectral and 3D data

Tropical forests contain incredibly high biodiversity but face multiple threats, contributing to the global biodiversity crisis. Studying tropical forest ecology is difficult because the sheer variety of co-existing species makes data collection challenging. Artificial intelligence (AI) promises to make large-scale, species-specific data collection on ecological function easier. AI can sort through vast quantities of data collected by tools like smartphones or remote sensing, identify species, and spot subtle differences in color or anatomy that can shed light on how the plants function within the ecosystem. In this project, researchers will collect hyperspectral and laser scanning data on over 150 tropical tree species in Panama, train AI to identify them, and link this information to long-term data on plant performance across environmental gradients. This AI approach and others like it should help researchers better understand and conserve biodiversity in species-rich forests around the world.