Tag Archives: Acoustics

Machine Learning and Acoustics

As computer systems continue to improve, there is an increased ability to complete tasks using artificial intelligence. A computer system can be trained to perceive its environment, make decisions, and take actions. One of the methods for achieving this is machine learning (ML), where machines ‘learn’ tasks from without the need for explicit programming. Given sufficient training, ML can process large, complex datasets to reveal patterns.

abstract blackboard bulb chalk

Photo by Pixabay on Pexels.com

Acoustics is one of the fields generating large amounts of data. For example, every time I deploy a hydrophone in the English Channel, it can record for 3 weeks and return approximately 250 GB of data! It doesn’t take long for the data to quickly add up to an overwhelming amount. Never mind the added challenge of searching through it for sounds of interest, such as fish and vessels. ML could provide the answer, if I could train the computer system to recognise these sounds and search the data for the presence. Similarly, ML could also greatly contribute to other types of acoustic research, like speech processing, sound localisation, noise mitigation, health monitoring, and so on.

But how do young acousticians (yes, I am still counting myself in this category!) learn how to develop and use ML tools? That is where the UK Acoustic Network (UKAN) comes in! The network contains several special interest groups, including one specially devoted to Early Career Researchers that holds an annual Summer School. This year, their aim was to provide training for early career acousticians to get to grips with ML tools. So I joined over 30 scientists at Gregynog Hall in Wales this week for the UKAN Summer School 2019 (#UKANSS19).

The week started with an introduction to Python. This was led by Dr Prasun Ray, a Strategic Teaching Fellow in Applied Mathematics at Imperial College London. Prasun took us through the fundamentals of programming and data analysis in Python, alternating between a mixture of seminars and interactive examples to put us through the paces. Although I haven’t used Python much since 2016, it was good to re-familiarise and get updated on what the software can do!

The ML classes were delivered by Dr Ramon Fuentes (Research Scientist at Callsign Ltd and Visiting Researcher at the University of Sheffield), who has previously applied ML for signal processing and the development of autonomous inspection systems. Although the mathematics of this went whooshing over my biological head, it was still interesting to learn some of the fundamental ideas that ML is built upon. Although anywhere Ramon speaks in the future might want to invest in some additional whiteboards…

Finally, we had a series of lectures on audio and speech applications of ML delivered by Professor Nilesh Madhu from Ghent University. In my opinion, these were the highlight of the week! Nilesh was an excellent speaker, expertly leading us from one topic to another with a strong storyline, a balanced level of detail, plenty of examples, and good humour. His own research focus is on signal detection, analysis and enhancement with application to mobile devices. Despite this seeming fairly different from my own work on the surface, it had a lot of deeper similarities. For example, the problem of recognising and classifying signals in challenging contexts will be particularly relevant to my “Decoding the Deep” project.

As well as generating ideas, the other great thing about this event was meeting people. All too often it is easy to get ‘stuck’ in your own research bubble, only speaking with or working alongside people in the same field. Whilst this can be a good thing, it is still good to step back every so often and look around – because that’s how ideas are born! UKAN exists to support and facilitate networking – and that was a strong highlight of this week. Living with 30+ acousticians for five days means that you never know when an interesting conversation will arise: over the breakfast table; during a lunch break; whilst trying out some casual archery; building a cardboard tower in a competitive team-building exercise; on a hike through the grounds; in a subterranean prison surrounded by the ghosts of past inmates… (not even joking). There were lots of opportunities for networking, bonding, and discussions throughout the week. I think everyone went home with some new contacts and friends, several of which I will be following up with over coming months.

Overall, a successful week for the UKAN Summer School 2019! The ECR Special Interest Group are planning to run another next year, along with some smaller events. Be sure to get in touch to register your interest, propose other events, or join the network!

 

SMM 2017 Conference Presentation

I’m now safely home after attending the Society for Marine Mammalogy’s biennial conference in Halifax, Nova Scotia.¬† It was an amazing eight days of conference talks and workshops, interspersed with plenty of networking.¬† As well as meeting lots of interesting new people, it was particularly wonderful to catch up with so many old friends.¬† Totally worth the 35hrs of one-way travel and 12hr time difference jetlag!

I’m currently preparing a detailed conference report for the Journal of Animal Ecology’s blog, but in the meantime I wanted to make my conference presentation available online.¬† To download a PDF version, click here.¬† The full abstract is provided below.¬† Feedback welcome!

SMM 2017 - Sarah Marley

Acoustic habitats and behavioural responses of bottlenose dolphins in Western Australia

Marley, S.A., Salgado Kent, C.P., Erbe, C., Parnum, I.M. and Parsons, M.J.G.

As human activities continue to expand across the marine environment, anthropogenic ocean noise is also rapidly increasing.¬† This is of concern to acoustically-specialised species, particularly those displaying a high degree of habitat overlap with anthropogenic activities, such as bottlenose dolphins (Tursiops sp.).¬† There is a need to describe the soundscape of coastal dolphin habitats and examine how prominent anthropogenic noise sources may impact these animals.¬† The Swan River in Western Australia flows through the state capital of Perth, containing over 1.4 million people, and consequently experiences a range of anthropogenic activities.¬† However, the river is also extensively used by a resident community of Indo-Pacific bottlenose dolphins (T. aduncus).¬† Autonomous underwater acoustic recorders were used to collect data throughout the Swan River, which were analysed via weekly spectrograms, power spectrum density percentile plots, octave-band levels, broadband noise levels, and generalised estimating equations.¬† Land-based theodolite tracking at two sites provided information on vessel traffic and dolphin behaviour, which were assessed using generalised additive models and Markov chains.¬† Acoustic datasets collected from 2005 to 2015 indicated that the Swan River was comprised of multiple acoustic habitats, each with its own characteristic soundscape and temporal patterns in underwater noise.¬† The ‚Äėnoisiest‚Äô site from an anthropogenic perspective and in relation to dolphin communications was the Fremantle Inner Harbour (mean broadband noise level: 106 dB re 1 ¬ĶPa rms [10 Hz ‚Äď 11 kHz]).¬† Theodolite tracking at this site found that dolphins remained present during periods of high vessel traffic.¬† However, behavioural observations indicated significant alterations to dolphin movement speeds and activity states at high vessel densities.¬† Furthermore, whistle characteristics varied in conditions of high broadband noise.¬† This work suggests that dolphins maintain occupancy at key foraging sites within the Swan River despite the presence of vessels, but alter their behaviour in periods of high vessel traffic.

New Paper: A Tale of Two Soundscapes

I am a little bit behind in posting about my various PhD publications.  But a pretty cool one that came out in the August edition of Acoustics Australia focused on comparing the acoustic characteristics of urban versus pristine coastal dolphin habitats.

I’ve previously spoken about the need to define the soundscape of dolphin habitats and examine patterns of when and where noise occurs.¬† But whilst we can monitor dolphin acoustic habitats and describe how they are¬†now, in many cases it is hard to say how they were in the¬†past.¬† We simply don’t have enough long-term datasets that capture the expansion of human activities into the marine environments – i.e. datasets that go far enough back in time to capture when the habitat was pristine and free of human influence.¬† Understandably, this can also make it hard to decide how we best manage man-made noise in these habitats into the¬†future.¬† It’s hard to determine what noise mitigation measures are necessary when you don’t know how quiet an area is ‘supposed’ to be.

A potential solution to this conundrum for managing ‘urban’ areas is to find comparable ‘pristine’ areas which are also used by the study species, and see what differences exist.¬† Of course, this is reliant upon pristine areas still existing in the first place.¬† However, in Western Australia we are fortunate enough to have some areas up in the remote north of the state.

The Kimberley region in north-western Australia is one of the most remote, pristine wildernesses in the world.¬† The largest town in the region is Broome, which has a permanent population of only 16,000 people (although the temporary population can increase to 45,000 in the tourism season).¬† Broome is located on the shores of Roebuck Bay, with a relatively large population of snubfin dolphins and is also regularly used by bottlenose dolphins and humpback dolphins.¬† This area offers a nice contrast with the busy, urban environment of the Swan River, which flows through the Western Australian state capital of Perth.¬† This river is also used by bottlenose dolphins.¬† To create successful noise mitigation strategies for coastal dolphins, there is a need to compare ‘quiet’ and ‘noisy’ acoustic habitats.¬† Roebuck Bay and the Swan River offer the perfect opportunity to do just this.

Prepping the Roebuck Bay acoustic gear

I chose to compare the pristine soundscape of Roebuck Bay with the Swan River’s anthropogenically-noisiest site, the Fremantle Inner Harbour.¬† I used autonomous underwater acoustic recorders to monitor the soundscape at these sites, and assessed these using a combination of weekly spectrograms, power spectrum density percentile plots and probability densities, octave-band levels, broadband noise levels, and generalised estimating equations – i.e. a shedload of nifty acoustic analyses.

What these essentially showed were that these two sites are very different in their acoustic characteristics.  In Roebuck Bay, biological sounds (such as crustaceans and fish) were the most prevalent sound sources, with very few instances of vessel noise.  However, in the Fremantle Inner Harbour, man-made noise dominated.  This worked out to a 20 dB difference between sites on average, and the frequencies used by dolphins for communication were more likely to be masked in Fremantle Inner Harbour based on elevated noise levels.

In this study, I was also lucky enough to get some acoustic recordings of sounds produced by snubfin and bottlenose dolphins in Roebuck, and could use these for a comparison with the sounds produced by bottlenose dolphins in Fremantle.  This helped me to discuss the potential consequences on Roebuck Bay dolphins if noise levels were to increase at that site.  Such information is going to be useful going into the future, as several coastal developments are currently planned for Roebuck Bay.

To find out more, check the paper out online or contact me!

Snubbies are happy little dolphins – hopefully they stay that way! (Photo: Sarah Marley)

 

New paper: Behavioural and acoustical responses of dolphins to vessel traffic and noise

Last night, my sixth (and final) PhD paper was published in the journal Scientific Reports!

In earlier papers, I established that the Swan River is a pretty noisy place.  One of the noisiest sites from an anthropogenic perspective is the Fremantle Inner Harbour, which experiences high levels of vessel traffic.  However, during my initial visual observations I discovered that dolphins continued using the harbour, even though it was a busy noisy environment.  So in this latest paper, I searched for evidence of dolphins responding to boats and noise at a subtler scale.

To do this, I used a combination of visual and acoustic monitoring techniques.  I found that dolphins significantly increased their movement speeds when vessel traffic occurred at high levels.  Similarly, dolphins also changed their behavioural budgets at high vessel densities, spending more time travelling and less time resting or socialising.

I monitored dolphin movement speeds and behaviours, to see how these changed as vessel traffic increased

I also looked for acoustic responses, by measuring nine characteristics to describe the shape and frequencies of dolphin whistles in different noise scenarios.  I found that all nine of these characteristics varied with increasing levels of broadband noise, and that the response was particularly strong for low-frequency noise.

I used nine characteristics to describe dolphin whistles

I’ve spoken before about the importance of scientific publishing, in terms of career brownie points for researchers.¬† So I’m pretty chuffed to have completed publishing my PhD!¬† To view the full list of papers from this and other projects, check out my Publications page.¬† Now to deal with all the ‘leftover’ PhD data that I collected, but didn’t get around to including!

New Paper: Spatial and temporal variation in dolphin acoustic habitats

There is a growing awareness of underwater noise in our oceans and the potential impacts of such noise on marine life, an issue which was the major theme of my PhD thesis. ¬†This is particularly relevant for “acoustically-specialised” species, such as dolphins. ¬†However, before we can start investigating the effects of noise on these animals, we first need to define the soundscape of dolphin habitats and examine patterns of when and where noise occurs.

We get a glimpse of them at the surface – but what about their acoustic habitat beneath the surface? (Photo: Sarah Marley)

This description of noise in dolphin habitats formed my third PhD chapter, which has now been published online in¬†the scientific journal¬†Frontiers in Marine Science. ¬†In this paper, I examine spatial and temporal variability in the soundscape of the Swan River using over 11,600 hours of acoustic data collected from five sites within the river system across eight years. ¬†Multiple sound sources were recorded at these sites, but the prevalence of these sounds at each site differed, giving each location a characteristic soundscape. ¬†Consequently, some sites were ‘noisier’ than others.

Deploying acoustic recording equipment with help from Fremantle Ports. Spot the dolphins in the background! (Photo: Jeanette Murray)

Want to know more? ¬†The full paper is available online! I’d love to hear your thoughts ūüôā

Marley et al. (2017) Spatial and temporal variation in the acoustic habitat of bottlenose dolphins (Tursiops aduncus) within a highly urbanised estuary.  Frontiers in Marine Science, 4: 197.

And watch this space, as the remaining three PhD papers should be going up online in the next few months!

 

 

New Video: The Dolphin and the Sound

As you know, I’m fortunate enough to be sister to the amazing wildlife film-maker Lisa Marley. ¬†I’ve written about her documentary on Scottish raptor poisonings here previously (and incidentally, this work is currently touring the film festivals¬†– scroll to the end for details!). ¬†But in June, we actually worked together on a short film project as part of the Aquatic Noise 2016 conference I attended in Dublin.

The conference held a public evening involving short lectures around the theme of underwater noise, and also invited submission of videos on this topic.  Lisa and I worked together to create a short film describing the effects of human noise on coastal dolphins, similar to the idea of my 3MT speech Рbut with much cooler visuals than just me standing on a stage!

I wonder if anyone recognises the locations involved in this production? ¬†Suggestions on a postcard please…!

So now I have a professional video to highlight my research AND had the awesome experience of working on a creative project with my sister.  And all without a single sibling squabble to be seen!


“Red Sky on the Black Isle” lastest screenings:

  • Hebrides International Film Festival (on Lewis, Harris, Uist and Barra; UK): ¬†14-17 September 2016
  • Aberdeen Film Festival (UK): ¬†17 October 2016
  • Festival de Menigoute (France): ¬†27 October 2016

Follow the film’s Facebook page for more updates!

Conference Attendance – Aquatic Noise 2016

Last month I had the privilege of flying to Dublin to attend¬†this year’s “Effects of Noise on Aquatic Life” conference, also known as Aquatic Noise 2016. ¬†Not only was I able to present the first two chapters of my PhD thesis, but I was able to contribute to the conference itself as part of the Media Committee.

Members of CMST attending the Aquatic Noise 2016 Conference in Dublin

Members of CMST attending the Aquatic Noise 2016 Conference in Dublin

Over 300 people from 23 countries attended the week-long conference, including representatives from universities, government research institutions, fisheries, and industry groups.  Talks were given regarding a variety of acoustic topics, including descriptions of noise sources, sound propagation analyses, and the responses of numerous animal species.  Conference attendees also included ten members of my lab, the Centre for Marine Science and Technology (CMST).  It was great to head overseas as a team, presenting our research to an international audience!

Poster PicMy talk was part of the student speed sessions – five minutes to describe our research and findings, followed by an evening poster session where we had the chance to answer questions, engage in discussions, and network. ¬†I feel the talk went well, despite my nerves. ¬†I’m much more used to interactive presentations involving members of the public than presenting to seasoned professional scientists. ¬†But even without any audience participation, dramatic displays or unexpected explosions, I think I did alright! ¬†You can view my poster by clicking here: ¬†Marley AN2016 Poster.

Marley Dolphin Presentation AN2016

Being on the Media Committee meant I was part of a team responsible for promoting the conference, particularly our public evening. ¬†This was attended by over 70 members of the public, who had an evening of lectures and short videos (more about this in my next post!). ¬†I was also involved in live-tweeting the conference. ¬†This was a way of extending the conference reach to the masses, scientists or otherwise, by posting summaries of each talk on Twitter under the conference¬†hashtag. ¬†I’ve now summarised each day of talks as a separate story using a combination of Tweets from myself and other conference attendees, which are available for anyone to read using the links below:

https://storify.com/sarahmarley86/aquatic-noise-2016-day-1

https://storify.com/sarahmarley86/an2016-effects-of-noise-on-behaviour-and-physiolog

https://storify.com/sarahmarley86/aquatic-noise-2016-day-3

https://storify.com/sarahmarley86/aquatic-noise-2016-day-4

https://storify.com/sarahmarley86/aquatic-noise-2016-day-5

New Paper: Fish Choruses in Darwin Harbour

ICES Journal ArticleAs well as working on my PhD, I’m also lucky enough to be involved in other projects at the Centre for Marine Science and Technology (CMST) from time-to-time. ¬†Now the results from one of these projects have been published in the ICES Journal of Marine Science.

A couple of years ago,¬†I was asked to review acoustic data from Darwin Harbour, in the Northern Territory of Australia. ¬†I’ve written before about the variety of sounds produced by fish, and in the Darwin data we found oodles of different fish choruses. ¬†Fish sounds can be species and size specific, and such en masse¬†sound production often has behavioural associations, for example by¬†corresponding with feeding or reproduction.

So there is a lot of information to be gained by listening in on fish!

In this paper, we recorded nine different types of fish choruses and investigate their patterns of occurrence. ¬†Environmental conditions such as lunar patterns, time of sunset, temperature, tidal information and salinity levels all contributed to the context of when particular choruses were heard. ¬†These results are¬†useful not only to scientists but also to fisheries managers, as it provides improved knowledge regarding¬†species distribution, fish habitat-use, identifies spawning seasons, and monitors behaviour. ¬†Which, when you can’t see fish below the surface, is often difficult data¬†to collect!

The full paper (doi: 10.1093/icesjms/fsw037) is available here.

First PhD Chapter Published!

After months of graft, I’m extremely chuffed to announce that my first PhD chapter has now been published online!

Chapter 1 - Marley et al 2016

To view the full article, please visit the Springer website.  If you have any issues, please contact me.

One down, one draft on supervisor’s desk, and three more to go!

Famous Last Words

FameLab Australia Final

MC Robyn Williams revs up the crowd at the FameLab Australia Final 2015 (Photo: OK White Lane)

My last post spoke of my experience competing in the WA State Heat of FameLab.  Although I was delighted to be voted the Peoples Choice by audience vote, unfortunately only the winner and runner-up of each state heat progress to the National Final.

Or so I thought…

But a couple of weeks after the State Heat, I received a phonecall from FameLab Australia organiser Chris Hodge, inviting me back to present again as the British Council Wildcard entrant.  So I made it to the Final by a fluke!

Sussing out the Competition

The FameLab Australia Final involved entrants from New South Wales, Victoria, ACT, Queensland and Western Australia.  A mix of PhD students, post-docs, and scientists-with-real-jobs we covered a range of subjects from coughing guinea pigs to spider behaviour, gut bacteria to brain function, biofuel to artificial intelligence.  A mysterious group of people at the best of times, especially when most of us had never met before!

I had the chance to suss out one of the competitors, David Farmer, on a radio interview with ABC Melbourne presenter Lindy Burns.  Ironically, when discussing the Australian competition, the station had managed to select two Scottish people to interview.  So this combined with dolphins, lasers and squishy brains made for some great banter in a pretty unique interview!

With competitors like that, it was obviously going to be some stiff competition…

Love and Science

Media trainer Malcolm Love (far left) with some of the FameLab Australia 2015 finalists

Media trainer Malcolm Love (far left) with some of the FameLab Australia 2015 finalists (Photo:  OK White Lane)

The lead-up to the National Final included two days of media training with science communication guru Malcolm Love.¬† Chief Trainer of FameLab International, Malcolm was originally a freelance journalist in South America before working for the BBC as a producer on features and documentaries for over 20 years.¬† He is now a specialist in the public engagement of science, giving lectures on the subject at the University of West England and providing training for a range of science-stakeholders, as well as hosting his own weekly radio show “Love and Science“.¬† So this guy knows what he’s talking about.

In the training, we covered a variety of topics including body language, story-telling and interview tips.¬† But one of the best things about it was interacting with people who love science communication.¬† Many scientists still hold onto a fear of presenting to the public, and worry about “dumbing down” their research or coming across as boring.¬† But all the participants were obviously people who were passionate about their research, and it is hard not to get swept up in that kind of enthusiasm!¬† So it was an awesome two days of being a science geek with other science geeks and discussing how to turn other people into science geeks too!

The Final

Nothing like a bad dolphin joke to kick-start your presentation...  (Photo:  OK White Lane)

Nothing like a bad dolphin joke to kick-start your presentation… (Photo: OK White Lane)

A sell-out event with over 200 people in attendance, the final was a bit more nerve-wracking than the state heat.¬† But I always tell presenters that you just have to try changing the nervous energy into excited energy, so when I stepped up to the spotlight I tried to remember my own advice.¬† Unfortunately, I still felt my performance lacked the right mix of enthusiasm – even as I was speaking, I knew it sounded over-rehersed.¬† So although the crowd laughed in the right spots and seemed keen, I knew it wouldn’t be a winning presentation.¬† But that’s okay – there will be others!

Science is a serious business - need to work on that intense expression!  (Photo:  Ok White Lane)

Science is a serious business – need to work on that intense expression! (Photo: Ok White Lane)

The overall winner of FameLab Australia 2015 was Dr Sandip Kamath, with Dr David Farmer coming a close second.  Sandip is studying shellfish allergies at James Cook University, and spoke of his ambition to help people overcome these reactions Рwith the help of Mr Pinchy the lobster, his side-kick slash prop.  David moved away from squishy brains and lasers to give a fascinating description of his research at the Florey Institute of Neuroscience, investigating the cough reflex and brainstem function.  To see photos from the night, check out the British Council Flickr Page.

And, of course, we all had fun celebrating at the FameLab after-party!¬† The official function was in the WA Maritime Museum, with some speeches and lots of well-wishers…¬† But the scientists and British Council crew headed out into Fremantle to celebrate afterwards!¬† After all, we’re twelve of the top young science communicators in the country!

So now, as Sandip flies to the UK to compete in the FameLab International Final at the Cheltenham Science Festival, for the rest of us it is back down to Earth.¬† I’m back in a world of fieldwork, marking student assignments, and desperately trying to finish the first scientific paper of my PhD.¬† I can see all my new sci comm pals talking about the same reality bump on Twitter.¬† But to be honest, getting back to research is quite exciting enough… for now!