This dataset contains approx. 180,000 XML annotation files assigning behaviours (9 overall behaviours), animal bounding boxes and species names to the content of every frame across the 500 chimpanzee and gorilla videos of the PanAfrican2019 dataset captured at 24 frames per seconds. Note that this dataset does not contain the video data itself, which is held by the PanAfrican Programme. Full details of the contents are provided in the readme.txt file. For associated code please check out When using this dataset please cite the accompanying paper "Visual Recognition of Great Ape Behaviours in the Wild" available at
Date made available9 Jun 2021
PublisherUniversity of Bristol


  • animal biometrics
  • computer vision
  • animal identification
  • great apes

Cite this