Artificial intelligence helps protect from Avian influenza

In the fight against the costly impacts of avian influenza, protecting poultry flocks from contact with wild birds is an important deterrent. However, the methods for effectively controlling wild birds are limited. In Georgia, researchers developed a novel method for detecting wild birds on farms, using night vision technology and artificial intelligence. 

To aid in monitoring wild bird activities near poultry farms, the team developed a deep-learning model to track and identify the birds. A night vision camera collected data on wild birds near the farms and found there were two principal species: the gadwall and brown thrashers. More than 6,000 images from the video collected were then used to train and test the model. 

The model was able to precisely track and identify the birds with 95% accuracy, providing a new way of developing strategies to prevent wild birds from transmitting highly contagious avian influenza to chicken and turkey flocks. 

University of Georgia Agricultural Experiment Station;University of Georgia Cooperative Extension | Project supported by private grants and contracts.