https://doi.org/10.25678/000DHT
You're currently viewing an old version of this dataset. To see the current version, click here.

Data for: Automated single species identification in camera trap images: architecture choice, training strategies, and the interpretation of performance metrics

Scripts for: Automated single species identification in camera trap images: architecture choice, training strategies, and the interpretation of performance metrics, Burkard et al. (2024) EcoEvoRxiv (https://doi.org/10.32942/X2VW4V). A pipeline developed to train, evaluate and apply computer vision models to recognize the grey heron (Ardea cinerea) in camera trap data. Methods are described in preprint Burkard et al. (2024). Models trained and compared are a classification algorithm (MobileNetV2) and an object detection algorithm (YOLOv5x6).

Data and Resources

Citation

This Data Package

RIS BibTeX Citeproc JSON Datacite XML

The associated article

Burkard, Y., Francazi, E., Lavender, E., Dubach, T., Wehrli, S., Brodesen, J., et al. (2024). Automated single species identification in camera trap images: architecture choice, training strategies, and the interpretation of performance metrics. https://doi.org/10.32942/x2vw4v
RIS BibTeX Citeproc JSON

Metadata

  Publication Data Package for:
Open Data Open Data
Author
  • Burkard, Yannick
  • Francazi, Emanuele
  • Lavender, Edward John Nicholas
  • Brodersen, Jakob
  • Volpi, Michele
  • Baity Jesi, Marco
  • Moor, Helen
Keywords Ardea cinerea,camera trap,monitoring,machine learning,image classification,object detection
Taxa (scientific names)
  • Ardea cinerea
Organisms (generic terms)
  • birds
  • Grey heron
Systems
  • river
  • stream
Timerange
  • 2017-01 TO 2017-08
Geographic Name(s)
  • Streams around lake Lucerne
Review Level general
Curator Moor, Helen
Contact Baity Jesi, Marco <marco.baityjesi@eawag.ch>
DOI 10.25678/000DHT