A Convolutional Neural Network for Automated Detection of Humpback Whale Song in a Diverse, Long-Term Passive Acoustic Dataset

Abstract

Passive acoustic monitoring is a well-established tool for researching the occurrence, movements, and ecology of a wide variety of marine mammal species. Advances in hardware and data collection have exponentially increased the volumes of passive acoustic data collected, such that discoveries are now limited by the time required to analyze rather than collect the data. In order to address this limitation, we trained a deep convolutional neural network (CNN) to identify humpback whale song in over 187,000 h of acoustic data collected at 13 different monitoring sites in the North Pacific over a 14-year period. The model successfully detected 75 s audio segments containing humpback song with an average precision of 0.97 and average area under the receiver operating characteristic curve (AUC-ROC) of 0.992. The model output was used to analyze spatial and temporal patterns of humpback song, corroborating known seasonal patterns in the Hawaiian and Mariana Islands, including occurrence at remote monitoring sites beyond well-studied aggregations, as well as novel discovery of humpback whale song at Kingman Reef, at 5∘ North latitude. This study demonstrates the ability of a CNN trained on a small dataset to generalize well to a highly variable signal type across a diverse range of recording and noise conditions. We demonstrate the utility of active learning approaches for creating high-quality models in specialized domains where annotations are rare. These results validate the feasibility of applying deep learning models to identify highly variable signals across broad spatial and temporal scales, enabling new discoveries through combining large datasets with cutting edge tools.

The Next Wave of Passive Acoustic Data Management: How Centralized Access Can Enhance Science

“…The soundscape of the American Samoa site is likely influenced by coral reef fish (e.g., Pomacentridae), snapping shrimp and humpback whales. All are well documented sound producers in the North Pacific (Mann and Lobel, 1998;Munger et al, 2012;Zang et al, 2016;Kaplan et al, 2018;Allen et al, 2021).…”

“…Efficiently extracting this critical information and comparing it to other datasets in the context of ecosystem-based research management is a Big Data challenge that traditional desktop processing methods cannot address (Marx, 2013;Bhadani and Jothimani, 2016). Machine learning and artificial intelligence are increasingly playing a role in research applications for PAM datasets (Shamir et al, 2014;Shiu et al, 2020;Allen et al, 2021). The curation, management, and dissemination of passive acoustic datasets is another Big Data challenge where progress is just beginning.…”

Characterizing the Long-Term, Wide-Band and Deep-Water Soundscape Off Hawai’i

“…Periods of humpback whale (Megaptera novaeangliae) singing were identified using a Convolutional Neural Network trained on this and other HARP data sets (Allen et al, 2021). The recordings were binned into 75-s intervals, and marked as either positive or negative for humpback song presence based on a threshold that gave a precision of 0.97 and recall of 0.93.…”

Sounding the Call for a Global Library of Underwater Biological Sounds

“…As a result, reported distributions of these species are being expanded. Even some of the great whales are being found in places they were not expected (Allen et al, 2021), and occasionally a new species (Rosel et al, 2021) or a new sound (Rice et al, 2014;Cerchio et al, 2020) is discovered. This fact could prove vital for soniferous fauna, as our ever-changing climate ensures that many species are modifying their distributions and broadening or reducing their ranges (e.g., Scheinin et al, 2011;Ramirez et al, 2017;Bonebrake et al, 2018).…”