The Brazilian Team is one of the six finalist teams in XPrize Rainforest, a five-year global scientific competition focused on diagnosing the biodiversity of tropical forests using innovative technologies, artificial intelligence and automation. The aim is to identify the largest number of species of flora and fauna in a 100-hectare area of rainforest.

Led by Prof. Vinicius Castro Souza, from Esalq/USP, the team brings together more than 50 researchers, mostly Brazilian, from various renowned institutions, such as USP, UFSCar and INPA. They use advanced technologies, such as autonomous robots and drones, to collect data on sounds, images, DNA samples, temperature, humidity and other environmental parameters, which are analyzed with artificial intelligence and other innovative techniques developed by Brazilian science.

The Brazilian Team’s proposal includes the use of deep learning to segment and identify Amazonian treetops. They carried out aerial mapping with drones over 1,700 hectares of the Amazon, where botanists identified more than 250 species of trees. With the images captured, it will be possible to train a deep learning model to identify the location of species throughout the mapped area, making it easier to identify new species and understand local biodiversity.