Beluga Whale Detection in the Cumberland Sound Bay using Convolutional Neural Networks

• Published in: Canadian Journal of Remote Sensing
• Main Authors: Lee, Peter Q., Radhakrishnan, Keerthijan, Clausi, David A., Scott, K. Andrea, Xu, Linlin, Marcoux, Marianne
• Format: Article in Journal/Newspaper
• Language: English
• Published: Taylor and Francis 2021
• Subjects: whale; marine; object detection; neural networks; Canada; Cumberland Sound; Beluga; Beluga whale; Beluga*; Béluga*
• Online Access: http://hdl.handle.net/10012/18113; https://doi.org/10.1080/07038992.2021.1901221

This is an Accepted Manuscript of an article published by Taylor & Francis in Canadian Journal of Remote Sensing on March 29, 2021, available online: https://www.tandfonline.com/doi/10.1080/07038992.2021.1901221 The Cumberland Sound Beluga is a threatened population of belugas and the assessment of the population is done by a manual review of aerial surveys. The time-consuming and labour-intensive nature of this job motivates the need for a computer automated process to monitor beluga populations. In this paper, we investigate convolutional neural networks to detect whether a section of an aerial survey image contains a beluga. We use data from the 2014 and 2017 aerial surveys of the Cumberland Sound, conducted by the Fisheries and Oceans Canada to simulate two scenarios: 1) when one annotates part of a survey and uses it to train a pipeline to annotate the remainder and 2) when one uses annotations from a survey to train a pipeline to annotate another survey from another time period. We experimented with a number of different architectures and found that an ensemble of 10 CNN models that leverage Squeeze-Excitation and Residual blocks performed best. We evaluated scenarios 1) and 2) by training on the 2014 and 2017 surveys respectively. In both scenarios, the performance on 1) is higher than 2) due to the uncontrolled variables in the scenes, such as weather and surface conditions. Natural Sciences and Engineering Research Council of Canada (NSERC), Grants RGPIN-2017-04869, DGDND-2017-00078, RGPAS-2017-50794, and RGPIN-2019-06744 || University of Waterloo || Marine Environmental Observation Prediction and Response Network