Wild Faces -Facial Recognition Software

Project Details

Principal Investigator: Prof. Dr. Sonia Kleindorfer

Funding: Australian Research Council

Programme: Linkage Project

Project Duration: 2022 - 2025

Project Number: LP210200740

Greylag Goose landing on the water

Picture: Josef Hemetsberger

Research Objective

We intend to fine-tune an automated identification software for individual identification of birds and then use facial recognition to obtain population, survival and demographic data for goose populations outside and inside their historic range. The proposed research aims at developing non-invasive monitoring tools of individual animals in the wild, with a focus on birds, to inform conservation priorities and set the standard for non-invasive monitoring by animal welfare principles. Nationally, the research will help researchers, government agencies and industry to track individual birds with minimal training and time to assess conservation efforts. We have three main aims:

1.     Develop a facial recognition software for individual identification in birds

2.     Assess the feasibility of facial recognition to obtain previously unobtainable population data in birds

3.     Determine the recovery of a vulnerable bird species following introduction outside of its natural range

Research Questions

The Greylag Goose (Anser anser) is a common and widespread goose species in Europe. We will field-test the individual facial recognition software in a flock of 140 individually colour-banded geese at the Konrad Lorenz Research Center in Austria. The flock of greylag geese was introduced into the Alm valley by the late Konrad Lorenz and co-workers in 1973. The geese are individually marked with coloured leg rings and are habituated to the close presence of humans. Individual life-history data have been collected since 1973 and provide reliable information about the age of the breeding birds. Data collection (e.g., photographs) across years from the same birds is possible and will provide further testing of the software accuracy across different time scales and seasons. We want to test the accuracy of identifying the correct goose using software and images of the same goose across time and in different conditions.



Greylag Goose, portrait from left side

Greylag Goose, portrait from left side

Greylag Goose, portrait from left side

Greylag Goose, portrait from left side

Pictures: J. Hemetsberger


KLF researchers and students

Dr. Diane Colombelli-Negrel, Flinders University

Prof. Dr. Tecumseh Fitch, University of Vienna

Prof. Dr. Leonida Fusani, University of Veterinary Medicine, Vienna


Prof. Dr. Sonia Kleindorfer

Phone: +43 (0)1 4277 76115

Email: Sonia.kleindorfer[at]univie.ac.at