Data augmentation for the classification of North Atlantic right whales upcalls.

Passive acoustic monitoring (PAM) is a useful technique for monitoring marine mammals. However, the quantity of data collected through PAM systems makes automated algorithms for detecting and classifying sounds essential. Deep learning algorithms have shown great promise in recent years, but their performance is limited by the lack of sufficient amounts of annotated data for training the algorithms. This work investigates the benefit of augmenting training datasets with synthetically generated samples when training a deep neural network for the classification of North Atlantic right whale (Eubalaena glacialis) upcalls. We apply two recently proposed augmentation techniques, SpecAugment and Mixup, and show that they improve the performance of our model considerably. The precision is increased from 86% to 90%, while the recall is increased from 88% to 93%. Finally, we demonstrate that these two methods yield a significant improvement in performance in a scenario of data scarcity, where few training samples are available. This demonstrates that data augmentation can reduce the annotation effort required to achieve a desirable performance threshold.

Performance of a deep neural network at detecting North Atlantic right whale upcalls.

Passive acoustics provides a powerful tool for monitoring the endangered North Atlantic right whale (Eubalaena glacialis), but robust detection algorithms are needed to handle diverse and variable acoustic conditions and differences in recording techniques and equipment. This paper investigates the potential of deep neural networks (DNNs) for addressing this need. ResNet, an architecture commonly used for image recognition, was trained to recognize the time-frequency representation of the characteristic North Atlantic right whale upcall. The network was trained on several thousand examples recorded at various locations in the Gulf of St. Lawrence in 2018 and 2019, using different equipment and deployment techniques. Used as a detection algorithm on fifty 30-min recordings from the years 2015-2017 containing over one thousand upcalls, the network achieved recalls up to 80% while maintaining a precision of 90%. Importantly, the performance of the network improved as more variance was introduced into the training dataset, whereas the opposite trend was observed using a conventional linear discriminant analysis approach. This study demonstrates that DNNs can be trained to identify North Atlantic right whale upcalls under diverse and variable conditions with a performance that compares favorably to that of existing algorithms.

A social and ecological assessment of tropical land uses at multiple scales: the Sustainable Amazon Network.

Science has a critical role to play in guiding more sustainable development trajectories. Here, we present the Sustainable Amazon Network (Rede Amazônia Sustentável, RAS): a multidisciplinary research initiative involving more than 30 partner organizations working to assess both social and ecological dimensions of land-use sustainability in eastern Brazilian Amazonia. The research approach adopted by RAS offers three advantages for addressing land-use sustainability problems: (i) the collection of synchronized and co-located ecological and socioeconomic data across broad gradients of past and present human use; (ii) a nested sampling design to aid comparison of ecological and socioeconomic conditions associated with different land uses across local, landscape and regional scales; and (iii) a strong engagement with a wide variety of actors and non-research institutions. Here, we elaborate on these key features, and identify the ways in which RAS can help in highlighting those problems in most urgent need of attention, and in guiding improvements in land-use sustainability in Amazonia and elsewhere in the tropics. We also discuss some of the practical lessons, limitations and realities faced during the development of the RAS initiative so far.