Case Study: Naturalis improves insect species identification with Capgemini's AI and AWS solution

Naturalis - Customer Case Study

Naturalis teamed up with Capgemini and AWS to tackle the challenge of protecting biodiversity by improving the identification of insect species from sound recordings. Through Naturalis’ involvement in Capgemini’s Global Data Science Challenge, the organization sought better machine learning methods to monitor insect populations, which are difficult to track but essential to healthy ecosystems.

Capgemini implemented a collaborative AI hackathon using AWS machine learning services, including Amazon SageMaker, to test models on five hours of audio from 66 insect species. The results were strong: macro-averaged F1 score improved from 76% to 92%, and Capgemini, Naturalis, and AWS published the outcomes so other organizations can build on them to advance insect conservation.

Naturalis

Elaine van Ommen Kloeke

Naturalis