discussion / Acoustics  / 10 March 2020

Virtual Meetup Discussion: Acoustic Monitoring

Hi everyone. 

We're a few hours away from our Acoustic Monitoring Virtual meetup. If you want to join us. there is still time to register here. We're at ~235 registrations from 45 countries, so it's going to be a popular one! 

From your feedback to our questions in registration, it looks like there are few broad topics that we're going to touch on in the discussion: 

How does acoustic monitoring compare to other techniques? - ‘results versus status quo

  • (UK): How advanced is Ecological Monitoring at this point in time and how is it advanced over last decade?

  • (Portugal): How different is acoustic data from other sampling techniques data (e.g. mist-nets vs acoustic detectors)?

  • (USA): How effective are the technologies? (results versus status quo)

  • (Australia): What can long duration acoustic datasets tell you than other kinds of monitoring cannot?

  • (UK): How easy would it be for someone with no prior knowledge of acoustic monitoring to successfully use it for monitoring a project?

Performance/Kit selection – has anyone done side-by-side comparisons of hardware?

  • (USA):  Has anyone done side-by-side comparisons of audiomoths and SM4s

  • (UK): I'm a physicist/engineer with an interest in acoustics trying to enter the field of conservation. Best devices for acoustic monitoring underwater? Store bought vs open source vs designing own monitoring devices?

  • (USA): I want to learn more about µMoth functionality and availability

Hardware: specific recommendations for kit and best practice. - NB: Given the detail in some of these questions (there a lot about marine applications, and specific set ups, e.g. around nests), they are probably going to be ones we suggest are best posted to the acoustic monitoring group for community knowledge exchange after the meetup

  • (UK): Interested in acoustic monitoring for migratory birds. Can the facilitators/participants recommend a very durable monitoring tool that can be left in remote areas for long periods?
  • (UK): Potentially considering use of acoustic monitoring on curlew nests. Distance from nests that equipment can be used successfully, and how easily concealed would it be?
  • (Australia): What are practical field techniques?
  • (USA):What are best practices for using recorders to survey bird biodiversity?
  • Marine Monitoring: (UK): Best devices for acoustic monitoring underwater? (Canada): Working on management of underwater noise in the Arctic. Framework for developing systematic acoustic monioring in the ocean.

Best way to analyse data? - there were a lot of questions here. Again, we'll cover some of these in the talks and discussion, but we're probably going to be suggesting folks jump into the acoustic monitoring group for in depth discussion and feedback. 

  • (UK): We conduct acoustic monitoring in Indonesia with devices we co-developed with partners from Invisible Flock (Art-Tech). Best way to analyse the sound files
  • (Democratic Republic of Congo): How to easily analyse a large amount of bat echolocation data?
  • (Canada): Is there a future for population size estimates or isolating individuals with acoustic analyses?
  • (Colombia): How to best measure and characterize the acoustic environment in order to address ecological questions?
  • (Norway): How to go about detecting damaged recordings in large sound datasets
  • (USA): Are there template schemas/best practices for database management?
  • (Malaysia): I am currently doing collecting bat echolocation calls for the Southeast Asian bat call database. How do I  minimize/reduce noise from my acoustics data datasets
  • (Nepal): How to analyse bat calls for species identification.
  • (UK): I work on animal monitoring sensors technologies. What strategies have already been implemented to identify individual animals based on their particular vocalisations?

Future: Predominantly about Machine Learning and Automatic ID - where are we at?

  • (USA): Our lab is involved in large scale terrestrial acoustic monitoring (mostly birds, some frogs) Opinions on "what's next" from a machine learning perspective

  • (Mexico): Are their automatic algorithms for ID (interested in birds)?

  •  

    (UK): Already doing acoustic monitoring of nocturnal birds. Find out more about latest on neural network architecture
  •  

    (USA): How effective are AudioMoths at monitoring bats? Is autoclassification of bird vocalizations on the horizon?

  • (USA): How easy to adopt software designed for birds to other taxa like frogs

  •  

     

     

    (Canada): Have you been able mesh network ARU's successfully? Is it something that you're actively working on? Why/why not?

  • (UK): What are the current knowledge gaps in the research?

  • (USA): What are current needs of bioacousticians for new technology?

  •  

     

     

     

     

    (USA): Are there turnkey field systems available? solar power? wireless? AI analysis/reporting? include pollinators?

  • (india): Field feasibility of using real time acoustic monitoring

How do we collaborate? Data sharing? Sound event detection libraries? Open Source?

  • (USA): Can we get more AI/ML students to work on projects like this before they go into industry?
  • (Uruguay): How do we create an audiomoth user community? Any users out there recording moth Ultrasounds?
  • (Canada): Are there state-of-the-art open-source libraries for sound event detection? (studies owl monkeys)
  • (Australia): How can we move forward with a template recognition database?
  •  (USA): We are a research project of future.edu, tracking climate change with bioacoustics in vulnerable Nepal, Bolivia, and US. Interested in conversations around vulnerable bird and amphibian species.  Collaborate on Tensorflow acoustic AI models?
  • (Spain): are there any open source projects that we can look at to get started?
  • (USA): what's the best source for open-access spectrogram datasets
  • (USA): What should we be doing as a community to build large open training datasets?
  • (Philippines): How to integrate citizen science to data collection and analysis?

 

See you all very soon!

Steph 




Hi all,

I missed this, but just watched the virtual meeting recording.  Towards the end Stephanie mentioned there were many questions about recording multiple channels for localisation.  I hadn't thought there was that much interest in it.  Anyway my collaborator/client and I have been working on this off and on and slowly, but she did present some results she got at the recent conference in NZ.  If there is any interest I'll see if I can get her to give out her slides or something.

On another tack I'm interested in the experience of anyone who has done gunshot/blast fishing localisation.  In particular I'm curious about the algo used for detection and also for localisation.  I'm taking the matched filter approach for detection but am now wondering if maybe CNN classification might be better.

Finally I'm interested if anyone has ideas about implementing a LoRa mesh, or knows of a useable mesh layer.

Thanks,

-harold