Work

Camera traps and deep learning enable efficient large‐scale density estimation of wildlife in temperate forest ecosystems

Year: 2025

Type: article

Abstract: Automated detectors such as camera traps allow the efficient collection of large amounts of data for the monitoring of animal populations, but data processing and classification are a major bottleneck. Deep learning algorithms have gained increasing attention in this context, as they have the potential to dramatically decrease the time and effort required to obtain population density estimates. However, the robustness of such an approach has not yet been evaluated across a wide range of species and study areas. This study evaluated the application of DeepFaune, an open‐source deep learning algorithm for the classification of European animal species and camera trap distance sampling (CTDS) to a year‐round dataset containing 895,019 manually classified photos from 10 protected areas across Germany. For all wild animal species and higher taxonomic groups on which DeepFaune was trained, the algorithm achieved an overall accuracy of 90%. The 95% confidence interval (CI) of the difference between the CTDS estimates based on manual and automated image classification contained zero for all species and seasons with a minimum sample size of 20 independent observations per study area, except for two. Meta‐regression revealed an average difference between the classification methods of −0.005 (95% CI: −0.205–0.196) animals/km 2 . Classification success correlated with the divergence of the population density estimates, but false negative and false positive detections had complex effects on the density estimates via different CTDS parameters. Therefore, metrics of classification performance alone are insufficient to assess the effect of deep learning classifiers on the population density estimation process, which should instead be followed through entirely for proper validation. In general, however, our results demonstrate that readily available deep learning algorithms can be used in largely unsupervised workflows for estimating population densities from camera trap data.

Source: Remote Sensing in Ecology and Conservation

DOI: 10.1002/rse2.70030