Data from: Drones and deep learning produce accurate and efficient monitoring of large-scale seabird colonies

Public

Population monitoring in some colonial seabirds is often complicated by the large size of colonies, remote locations, and by close inter- and intra-species aggregation. While drones have been successfully used to monitor large inaccessible colonies, the vast amount of imagery collected introduces a data analysis bottleneck. Convolutional neural networks are evolving as a prominent means for object detection and can be applied to drone imagery for population monitoring. In this study, we explore the use of these technologies to increase capabilities for seabird monitoring by using convolutional neural networks to detect and enumerate Black-browed Albatross (Thalassarche melanophris) and Southern Rockhopper Penguins (Eudyptes c.chrysocome) at one of their largest breeding colonies, the Falkland (Malvinas) Islands. Our results show these techniques have great potential for seabird monitoring at significant and spatially complex colonies, producing accuracies of 97.66% and 87.16% with ninety percent of automated counts being within 5% of manual counts. The results of this study imply our methods are a viable population assessment tool, presenting many opportunities to reduce manual labor, cost, and human error.

This dataset contains all relevant drone imagery and associated training, testing, and validation labels for the creation of a convolutional neural network to detect seabirds. ... [Read More]

Data Citation

Hayes, M. C., Gray, P. C., Harris, G., Sedgwick, W. C., Crawford, V. D., Chazal, N., Crofts, S., & Johnston, D. W. (2020). Data from: Drones and deep learning produce accurate and efficient monitoring of large-scale seabird colonies. Duke Research Data Repository. https://doi.org/10.7924/r4dn45v9g

Creator: Sedgwick, Wade C.
Johnston, David W.
Gray, Patrick C.
Crofts, Sarah
Hayes, Madeline C.
Harris, Guillermo
Chazal, Natalie
Crawford, Vivon D.
DOI: 10.7924/r4dn45v9g
Subject: Deep learning
Drone aircraft
Convolutional neural network
Neural networks (Computer science)
Sea birds
Publication Date: September 17, 2020
ARK: ark:/87924/r4dn45v9g
Publisher: Duke Research Data Repository
Type: Dataset
Related Materials: Main code repository: https://github.com/madelinehayes/seabirdNET
Training, validation, and inference code: https://github.com/fizyr/keras-retinanet
Hayes, M. C., Gray, P. C., Harris, G., Sedgwick, W. C., Crawford, V. D., Chazal, N., Crofts, S, & Johnston, D. (2021). "Drones and deep learning produce accurate and efficient monitoring of large-scale seabird colonies," Ornithological Applications, 123(3), 1-16. https://doi.org/10.1093/ornithapp/duab022
Funding Agency: Wildlife Conservation Society and the Island Foundation
Contact: Madeline Hayes: madeline.c.hayes@duke.edu
License: Creative Commons CC0 1.0 Universal
Title: Data from: Drones and deep learning produce accurate and efficient monitoring of large-scale seabird colonies
License: http://creativecommons.org/publicdomain/zero/1.0/
Repository: Duke Research Data Repository

Title	Date Uploaded	Visibility	Actions
HayesM_README.txt	2020-12-16	Public	Download
Training, Validation, and Testing Labels and Tiles	2020-09-17	Public
Drone Imagery	2020-09-17	Public

Usage Stats