Data and code release of Paying Attention to Other Animal Detections Improves Camera Trap Classification Models

Data and code release of: Paying Attention to Other Animal Detections Improves Camera Trap Classification Models

 

Paper in open-access (Methods in Ecology and Evolution) : https://onlinelibrary.wiley…

 

Code is available on GitHub : https://github.com/gdussert… and also archived in MCA_Classifier-code.zip

 

This repository contains:

 

* models:

   * crop_classifier.pt

   * mca_classifier.pt

* data:

   * serengeti_train: training set of Snapshot Serengeti

   * serengeti_test: test set of Snapshot Serengeti

   * serengeti_toy: toy dataset to quickly test all the scripts, it contains only the pictures of Snapshot Serengeti camera trap location C12

   * safari2024: whole Snapshot Safari 2024 Expansion dataset

   * crop_images.zip: crop images of Snapshot Serengeti camera trap location C12

   * speciesnet.zip: SpeciesNet predictions for all images, as well as the script used to compute the accuracy reported in the manuscript. 

* code:

  * MCA_Classifier-code.zip

 

Please cite our paper for the models: 

 

@article{dussert_paying_2026,
  author = {Dussert, Gaspard and Dray, Stéphane and Chamaillé-Jammes, Simon and Miele, Vincent},
  title = {Paying attention to other animal detections improves camera trap classification models},
  year = {2026},
  url = {onlinelibrary.wiley.com,
  doi = {10.1111/2041-210x.70260},
  journal = {Methods in Ecology and Evolution},
  publisher={Wiley Online Library}
}

 

Please cite this paper for the Serengeti dataset: 

 

@article{swanson2015snapshot,
  title={Snapshot Serengeti, high-frequency annotated camera trap images of 40 mammalian species in an African savanna},
  author={Swanson, Alexandra and Kosmala, Margaret and Lintott, Chris and Simpson, Robert and Smith, Arfon and Packer, Craig},
  journal={Scientific data},
  volume={2},
  number={1},
  pages={1--14},
  year={2015},
  publisher={Nature Publishing Group}
}