Related Experiment Video
Updated: Jun 29, 2025
05:16
Flying Insect Detection and Classification with Inexpensive Sensors
Published on: October 15, 2014
25.2K
VespAI: a deep learning-based system for the detection of invasive hornets.
Thomas A O'Shea-Wheller1, Andrew Corbett2, Juliet L Osborne3
1Environment and Sustainability Institute, University of Exeter, Penryn, Cornwall, TR109FE, UK. t.a.oshea-wheller@exeter.ac.uk.
Communications Biology
|April 3, 2024
Summary
An AI system called VespAI rapidly detects the invasive hornet Vespa velutina nigrithorax using a specialized monitoring station. This technology offers a robust early warning system to manage hornet populations and protect pollinators.
Area of Science:
- Ecology
- Artificial Intelligence
- Invasive Species Management
Background:
- The invasive hornet Vespa velutina nigrithorax poses a significant threat to European and East Asian pollinators.
- Early detection and eradication of V. velutina colonies are crucial for managing invasion spread.
- Current reliance on public visual alerts for detection is inaccurate and inefficient.
Purpose of the Study:
- To develop an automated system for the rapid and accurate detection of Vespa velutina.
- To leverage deep learning and AI for real-time invasive species monitoring.
- To provide a scalable solution for early warning and management of hornet invasions.
Main Methods:
- Development of VespAI, a hardware-assisted AI system.
- Utilizing a standardized monitoring station with YOLOv5s architecture and a ResNet backbone.
- Training a bespoke end-to-end pipeline for hornet detection.
Main Results:
- VespAI achieved real-time detection of V. velutina with a mean precision-recall score of ≥0.99.
- The system successfully sent image alerts via a compact remote processor.
- A prototype system demonstrated effective field operation and suitability for large-scale deployment.
Conclusions:
- VespAI offers a transformative approach to invasive hornet management.
- The system provides a robust early warning mechanism to prevent regional ingressions.
- Automated AI-driven detection enhances the efficiency and accuracy of invasive species control.
