One of the world’s most endangered whale species could have added protectionfrom threats posed by human marine activity, through technology developed bythe University of East Anglia (UEA).
In partnership with the Scottish Association for Marine Science (SAMS) and themarine survey company Gardline Geosurvey Limited, UEA researchers havedeveloped machine learning techniques that can be used to detect the presenceof North Atlantic right whales by listening for the sounds they makeunderwater.
Detecting the animals’ presence before they reach close proximity to largevessels or enter a mitigation zone can both protect animals and avoid costlyshutdowns of offshore operations.
The findings, ‘Robust North Atlantic right whale detection using deep learningmodels for denoising’, is published today in a special edition on machinelearning in acoustics, in The Journal of the Acoustical Society of America.
North Atlantic right whales are one of the world’s most endangered marinespecies with only around 350 remaining and of those, only about 100 females ofbreeding age. Human activities are a significant threat to right whalepopulations, either through entanglement in fishing gear or strikes fromshipping.
Calls from right whales are often confused with noises made from shipping orother underwater activities, such as fishing and drilling. The new techniquesdeveloped by UEA and its partners can remove these unwanted noises fromrecordings, thereby increasing the reliability of detecting right whales inadverse conditions.
The conventional way of locating right whales relies on observers onboardships, but this is expensive and not possible at night or in low-visibilityconditions. An automated method to detect the presence of right whales givesmuch more hope for the species to survive and increase in population, saidlead researcher Dr Ben Milner of UEA’s School of Computing Sciences.
Dr Milner, a senior lecturer, said: “The aim of this work is to develop robustmethods of detecting marine mammals from passive acoustic monitoring (PAM)devices in challenging environments.
“Having the ability to deploy an automated system – whether it be on buoys,Autonomous Surface Vehicles (ASVs), or gliders – that can achieve high levelsof detection in real-time, is vital to the long-term future of right whales.
“Being able to reliably detect marine mammals is important for populationmonitoring and for mitigation, as many species are endangered and protected byenvironmental laws.”
The technology aims to find right whales in situations where they could beapproaching potentially harmful and noisy offshore activities. In suchscenarios shipping can be asked to change course and in extreme situations theoffshore activities must be stopped, which can be very costly to operators.
Right whales emit a range of vocalisations, with common sounds being upcalltones and gunshot sounds. Upcalls most likely play a role as a social contactcall between individuals and are produced by both sexes and different ageclasses, and are therefore most commonly used for passive acoustic detectionof the species. The gunshot sounds are very different from upcalls and arecharacterised as an impulse, and although less common, can also be detected bythe new technology.
Both vocalisation types can be difficult to hear in noisy conditions and tovisualise in spectrograms, as low frequency regions are often masked by marinenoise from passing ships, drilling and piling activities, seismic exploration,or interference from other marine mammals, such as humpback whale song. Inmany cases, anthropogenic and environmental noises overlap in frequency withright whale calls, which makes detection difficult.
The researchers studied ‘de-noising’ processes that could block out non-whalenoises from trawlers, tankers and other human activities.
The right whale recordings used for evaluating the classifiers and denoisingmethods were taken from the Detection, Classification, Localization, andDensity Estimation (DCLDE) 2013 workshop and were collected in the Gerry E.Studds Stellwagen Bank National Marine Sanctuary from the Massachusetts Bayarea of the north-eastern coast of the United States.
‘Robust North Atlantic right whale detection using deep learning models fordenoising’, is published on 3 June 2021 in The Journal of the AcousticalSociety of America.
Previous Carp Diem: Pest seizes on post flood conditions, native fish at risk
Next In the USA Petco announces the expansion of its same-day delivery offresh pet food