LEVERAGING SOCIAL MEDIA AND DEEP LEARNING TO DETECT RARE MEGAFAUNA IN VIDEO SURVEYS.

Mannocci L, Villon S, Chaumont M, Guellati N, Mouquet N, Iovan C, Vigliola L & Mouillot D (2021).

Conservation Biology, 11, doi:10.1111/cobi.13798

Key message : We designed a method that takes advantage of videos accumulated on social media for training deep-learning models to detect rare megafauna species in the field. We trained convolutional neural networks (CNNs) with social media images of areial images of Dugons and tested them on images collected from field surveys. Our results highlight how and the extent to which images collected on social media can offer a solid basis for training deep-learning models for rare megafauna detection and that the incorporation of a few images from the study site further boosts detection accuracy. Our method provides a new generation of deep-learning models that can be used to rapidly and accurately process field video surveys for the monitoring of rare megafauna.

Steps in the deep-learning method that detects rare megafauna (CNN, convolutional neural network; TP, true positive; FP, false positive; FN, false negative). Social media images (gray) are used in steps 2 and 3. Field images (black) are used in steps 2, and 4.

Back to Nicolas Mouquet homepage ...

OTHER TOPICS: Aesthetics of Biodiversity, Biogeography, Macroecology & Ecophylogenetics, Experimental Evolution, Functional Biogeography, Functional Rarity, Nature for Future, Metacommunities, Metaecosystems, Reviews and Synthesis, Trophic Biogeography & Metaweb