Developing metadata standards for biodiversity in Bergen

What an adventure it has been to join the EU InsectAI project writing retreat 23 - 25 Sept 2024 in Bergen, Norway!

Under the umbrella of COST Action 22129 InsectAI: Using Image-based AI for Insect Monitoring and Conservation, researchers and practitioners from across Europe gathered to tackle a key challenge: developing guidelines for metadata standards for insect monitoring.

Why does this matter? Insect populations are reported to be declining globally, and monitoring them at scale is one of the greatest challenges for biodiversity science. Camera traps and image-based AI are powerful tools to document insect diversity, but without shared standards for metadata, that is, the information describing images and machine learning outputs, data remains fragmented and hard to reuse. This is an issue because decision-making across key stakeholder groups should be data-driven.

In Bergen, we got together to start shaping a framework that will help ensure that camera trap images carry consistent and useful contextual data, harmonise machine learning outputs so results are comparable, and lay the foundation for open and interoperable biodiversity datasets.

COST (European Cooperation in Science and Technology) makes this possible by funding networks that connect researchers, helping them exchange knowledge and grow new ideas together. You can learn more about COST Actions here.

A big thank you to the organisers and the entire @InsectAIers community for the energy, collaboration, and inspiration.