Seeing biodiversity: perspectives in machine learning for wildlife conservation

Tuia, Devis; Kellenberger, Benjamin; Beery, Sara; Costelloe, Blair R.; Zuffi, Silvia; Risse, Benjamin; Mathis, Alexander; Mathis, Mackenzie W.; van Langevelde, Frank; Burghardt, Tilo; Kays, Roland; Klinck, Holger; Wikelski, Martin; Couzin, Iain D.; van Horn, Grant; Crofoot, Margaret C.; Stewart, Charles V.; Berger-Wolf, Tanya

doi:10.48550/arxiv.2110.12951

Cited by 2 publications

(3 citation statements)

References 90 publications

(114 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our work also highlights that, when using sequence-based identification, the performance of classification models trained without sequence information are likely to be underestimated in contrast to what can be achieved at the sequence level. In the future, sequence or temporal information could be directly integrated into the training step, as done in context CNN models (Beery et al, 2020;Beery et al, 2021;Tuia et al, 2021). We could not use this approach here as many partners provided data without information about the specific trapping sites where the pictures came from.…”

Section: Model Performance and Relevance For Ecological Studiesmentioning

confidence: 99%

“…The full potential of camera-traps is however only realized when the hundreds of thousands of images, many being empty from spurious detections, can be rapidly classified with minimal human intervention (Chen et al, 2014;Schneider et al, 2019;Wearn et al, 2019;Tuia et al, 2021). Since the beginning, machine learning approaches, and in particular deep learning methods, have held the promise to solve this issue.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

The DeepFaune initiative: a collaborative effort towards the automatic identification of French fauna in camera-trap images

Rigoudy

Benyoub²,

Besnard

et al. 2022

Preprint

View full text Add to dashboard Cite

Camera-traps have revolutionized the way ecologists monitor biodiversity and population abundances. Their full potential is however only realized when the hundreds of thousands of images collected can be rapidly classified with minimal human intervention. Machine learning approaches, and in particular deep learning methods, have allowed extraordinary progress towards this end. Trained classification models remain rare however, and for instance are only emerging for the European fauna. This can be explained by the technical expertise they require but also by the limited availability of large datasets of annotated pictures, which are key to obtaining successful recognition models.In this context, we set-up the DeepFaune initiative (https://deepfaune.cnrs.fr), a large-scale collaboration between dozens of partners involved in research, conservation and management of wildlife in France. The aim of DeepFaune is to aggregate individual datasets of annotated pictures to train species classification models based on convolutional neural networks, an established deeplearning approach.Here we report on our first milestone, a two-step pipeline built upon the MegaDetector algorithm for detection (discarding empty pictures and cropping the animal) and a classification model for 18 species or higher-level taxa as well as people and vehicles. The classification model achieved 92% validation accuracy and showed > 90% sensitivity and specificity for many classes. Most importantly, these performances were generally conserved when tested on an independent out-of-sample dataset. In addition, we developed a cross-platform graphical-user-interface that allows running the pipeline on images stored locally on a personal computer.In conclusion, the DeepFaune initiative provides a freely available (for non-commercial purposes) toolbox with high performance to classify the French fauna in camera-trap images.

show abstract

Section: Model Performance and Relevance For Ecological Studiesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

The DeepFaune initiative: a collaborative effort towards the automatic identification of French fauna in camera-trap images

Rigoudy

Benyoub²,

Besnard

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Modern semantic image classification frameworks are almost without exception grounded in feedforward CNNs, introduced in their earliest form by the ground-breaking work of Krinzhevsky et al [23] and leading to further milestone architectures including VGG16 [41], Inception [43], ResNet [14], ResNeXt [49], and deep transformer networks [46]. Taxonomic computer vision (CV) applications of these techniques for microfossils are still very rare in the literature, despite steep advances in general animal biometrics [25,45]. Nevertheless, ML-based species identification [8,32] has been applied to several microscopic taxa, including coccoliths [42], pollen [44], and phytoplankton [48].…”

Section: Introductionmentioning

confidence: 99%

Visual Microfossil Identification via Deep Metric Learning

Karaderi¹,

Burghardt²,

Hsiang³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

We apply deep metric learning for the first time to the problem of classifying planktic foraminifer shells on microscopic images. This species recognition task is an important information source and scientific pillar for reconstructing past climates. All foraminifer CNN recognition pipelines in the literature produce black-box classifiers that lack visualisation options for human experts and cannot be applied to open set problems. Here, we benchmark metric learning against these pipelines, produce the first scientific visualisation of the phenotypic planktic foraminifer morphology space, and demonstrate that metric learning can be used to cluster species unseen during training. We show that metric learning outperforms all published CNN-based state-of-the-art benchmarks in this domain. We evaluate our approach on the 34,640 expert-annotated images of the Endless Forams public library of 35 modern planktic foraminifera species. Our results on this data show leading 92% accuracy (at 0.84 F1-score) in reproducing expert labels on withheld test data, and 66.5% accuracy (at 0.70 F1-score) when clustering species never encountered in training. We conclude that metric learning is highly effective for this domain and serves as an important tool towards expert-in-the-loop automation of microfossil identification. Key code, network weights, and data splits are published with this paper for full reproducibility.

show abstract

Seeing biodiversity: perspectives in machine learning for wildlife conservation

Cited by 2 publications

References 90 publications

The DeepFaune initiative: a collaborative effort towards the automatic identification of French fauna in camera-trap images

The DeepFaune initiative: a collaborative effort towards the automatic identification of French fauna in camera-trap images

Visual Microfossil Identification via Deep Metric Learning

Contact Info

Product

Resources

About