WG4 Data Analysis and Integration
Working Group 4 (WG4) is dedicated to addressing the downstream analysis of AI output to understand widespread insect declines using new technologies. The focus is on developing statistical methods for analyzing AI data and integrating it with traditional biodiversity data, alongside training ecologists in interpreting AI output. The objectives include establishing an analytical toolbox, connecting researchers across technologies, and training ecologists. Tasks involve developing statistical workflows and a web application to visualize biases in AI data, creating opportunities for parallel deployment of automated monitoring technologies, and developing a reproducible analytical workflow through Training Schools. WG4’s activities are tightly integrated with other working groups to ensure robust and transparent AI-based inference and address biases in insect monitoring methodologies.
Working group leads:
Dr Joseph Chipperfield
Norwegian Institute for Nature Research – NINA
Norway
joseph.chipperfield@nina.no
Dr Charlie Outhwaite
Institute of Zoology, Zoological Society of London
United Kingdom
charlotte.outhwaite@ioz.ac.uk
Dr Julie Koch Sheard
Philipps-Universität Marburg
Germany
sheard@staff.uni-marburg.de
Ongoing projects:
PAPER (in draft): The future of insect monitoring: integrating traditional monitoring practices with computer vision-based methods for robust trend analysis.
TOOLBOX (in progress): Awesome Analysis Tools
One of the key deliverables of Working Group 4 is a toolbox to support researchers working with insect monitoring data sourced from both traditional and computer vision-based collection methods. Monitoring data comes with a set of nuances depending on how and why it was collected. Understanding these nuances is essential for the interpretation of and inference from these data. This toolbox will link to existing tools, supply new tools, and offer guidance and support for researchers to appropriately assess and analyse their datasets.
STSM projects (looking for candidates): Please see the STSM page for current proposals.
