WWF's Conservation Innovation Awards

Submitted
Gathering votes
Support
Gathering votes
Originality
Refine
Close
Approved
This idea has great support - now the author needs to show that they have taken on board feedback, and scoped other ideas for potential collaboration. They need to consider this and answer additional questions which will show up in their idea post.
FINALIST

AviaNZ – An app to make bird identification easier

Birds are one of nature's joys, and many sections of the New Zealand community engage in some form of bird watching, from feeding garden birds to visiting our island safe havens. Birders would like to have an easy to access tool for identifying the birds that they hear. The rise of smartphone technology has enabled apps that perform this task for music, and the appeal of such an app for birdsong is clear. We focus to develop an app that:

  • Does not require any further information about the bird such as images, size, colour, or where the bird is, but simply an instant recording of the bird made by the phone itself.
  • Is more practical when the preferred species is nocturnal or hard to see.
  • Is ready to use - does not require tuning/ training.

 

What New Zealand conservation problem are you trying to solve?

Management of birds is essential to ensure their survival. This app helps to

  • Increase public participation in bird monitoring
  • Provide a learning platform
  • Decide presence/absence of a species in a forest
  • Understand the overall health of a forest
  • Build a recogniser for processing auto-recordings

 

What makes your idea new and unique?

Recently we developed a successful noise reduction method for natural bird recordings that usually carry diverse noise. Noise is one of the biggest challenges in automated bird recognition. The proposed app will use our noise reduction technique as well as machine learning techniques to accomplish the recognition.

Who will use your idea, and how will it make conservation action more efficient or effective?

Key users are general public in New Zealand and community groups.

What tasks or activities do you need investment for ? How would you spend a $25,000 grant?

  • Salary for a programmer
  • Purchase two phones: one Android and one iPhone for development and testing purposes

Are you a New Zealand citizen or resident?

Yes

Which category best describes your idea?

Research

I have read the Terms of Use, the Supplemental Terms of Use, and Community Guidelines - see links at the bottom of this page

Yes

List five other ideas posted in the challenge that excite you. Why?

DistaNZ: It solves a key problem in an environmental friendly way and it’s going to be accessible to community. So, it would be effective


Kill Trap: It's solving a burning problem in NZ


Catch-IT: This is a good e.g. of how we can use IT to make conservation easier


Lure, Trap & Retreat: It solves a similar problem to ours; the theme of both is the same but we focus on different domains and methods. In ours, birds are lucky because we don’t want to trap them to estimate their populations, but we capture their fabulous songs


Maximising the genetic diversity: It’s useful to save birds from extinction

How could you improve your idea?

 

Our idea has a great support from community and experts in the field. We confirmed having users from recreational bird watchers to serious bird monitoring experts and conservationists. Conservation efforts rely heavily upon acoustic methods to monitor population and ecosystem health for many of cryptic species in NZ. Auto recorders are rapidly being adopted for such surveys, yet most rely upon manual processing of recordings, which severely limits their efficacy. We have a great demand from community groups and DoC.  So, as this app is being developed we will get their feedback to parallel to developing a recogniser that is capable to process recordings made by any auto recorder. Our goal is an app for everyone and a more sophisticated birdsong recognition system for conservationists.

edited on Sep 23, 2015 by Nirosha Priyadarshani

Amal Punchihewa Sep 18, 2015

With ever increasing processing power and connectivity, it is important to explore the possibilities of developing robust algorithms that can perform (run) on portable devices. App (a computer program) that analyse and predict the bird with high accuracy s ia great idea to explore. 

 

Reply 2

Nirosha Priyadarshani Sep 21, 2015

Thanks for the very first vote for the idea. Yes, I remember our discussions on these things years ago... And I'm so glad, now we are exploring and finding exciting results!

Reply 0

Natasha Bansal Sep 20, 2015

Hi Nirosha, 

This app will be great! Not only will it get the community involved into wild bird monitoring but also help conservation and reduce costs. Great idea! Keep it going! 

Reply 2

Nirosha Priyadarshani Sep 21, 2015

Thanks Natasha! Of course that's our ultimate goal of this research.

Reply 0

Tim Arnold Sep 20, 2015

Hi Nirosha,

this is a great idea. Not only can many people use your idea and your idea enables projects to use crowd-sourced data, but it also is able to include the observation effort, by logging time and GPS tracks while on "survey". The effort is a crucial variable, essential to accurate abundance estimates, because you hear (or see) more if you seek longer. Your idea provides comparable data.

 

Reply 2

Nirosha Priyadarshani Sep 21, 2015

Thanks Tim. Yes, it can be used in different forms in different domains and can be certainly improved with extra information. It is true that in population estimations as well as identifying the presence/ absence of species, sampling could lead to less accurate results. For example, in presence/ absence monitoring of kiwi in a suspected forest, there are examples that observers only heard kiwi calling on the 14th or 15th consecutive night, but birds were silent for two weeks due to whatever reason... If we observe longer, our decisions are more accurate. Again the capacity of the observers also matters but programmable recorders can do the job smartly provided that there is an unbiased and accurate birdsong identifier to scan hours/ days/ weeks worth recordings without a pain. So we believe that in near future we can help with a tool that solves this practical problem in bird monitoring :)

Reply 1

Ellen Schoener Sep 20, 2015

You are doing great work Nirosha! This will make bird census easier, cheaper and faster, and therefore more can and will be done. 

Reply 2

Nirosha Priyadarshani Sep 21, 2015

Thanks a lot. Indeed this idea can save a lot and those will be invested to solve other conservation problems.

Reply 1

Rohana Ulluwishewa Sep 21, 2015

Hi Nirosha,

This is really a great idea to explore. I believe the app you are going to develop will be a great breakthrough in commutation technology as well as in environmental conservation.  The capacity of the proposed app to identify birds instantly at a relatively low cost makes your proposed project highly beneficial.   

Reply 2

Nirosha Priyadarshani Sep 21, 2015

Thanks for your support!

Reply 0

Richard Witehira Sep 21, 2015

Kia ora Nirosha,

Great idea, it would be amazing best of luck

Reply 1

Nirosha Priyadarshani Sep 21, 2015

Aku mihi nui ki a koe!

Reply 1

Isabel Castro Sep 21, 2015

Nirosha, as you know I am amazed by the great work you have done with the denoising, and very happy to see that you are now planning to use the denoising plus some further technology for this app. Conservation rests on knowing our natural world and what better way of learning than using an app such as the one you propose! 

Reply 1

Nirosha Priyadarshani Sep 21, 2015

So glad to hear your voice here! yes I'm halfway through and working to complete the mission :) This will be a great support for that...

Reply 1

Lorena Cruz-Bernate Sep 21, 2015

Hi Nirosha

AviaNZ is a great idea, congratulations and my best wishes for you!

Reply 1

Nirosha Priyadarshani Sep 21, 2015

Thanks Lorena :)

Reply 1

Tim Arnold Oct 2, 2015

*&(*#&*(#%& Bloody laptop touchpad, that was supposed to be a vote up!!!!

Reply 0

View all replies (2)

Nirosha Priyadarshani Sep 21, 2015

Thanks for every one who supported the idea along the way...

 

Reply 1

Edward Abraham Sep 21, 2015

Hi Nirosha, this is an excellent idea. An app like this has the potential to engage people with the natural world around them, and this is the key to conservation. The technical challenges that you solve in developing the app will also have broader application in the area of monitoring birds through their song. There is much promise with acoustic monitoring .

Reply 2

Nirosha Priyadarshani Sep 23, 2015

You are right Edward. Basically we want to make people more aware about birds around them with this app, especially the young generation in NZ. So we consider birds that we often hear as well as threatened/ endangered species. Our approach is not going to be species specific or even bird specific, that means we can use this to monitor introduced predators that are responsible for the endangered status of many bird and reptile species in NZ.

Reply 1

Pramila Gamage Sep 22, 2015

Hi Nirosha

This is a great idea...Congratulations and my best wishes for you....

Reply 2

Lee Barry Sep 23, 2015

Status changed to Finalist
This idea has passed all the milestones and is eligible for final judging by WWF's expert panel. But feel free to keep the comments coming in!

Reply 1

John Hart Sep 25, 2015

This is a great idea Nirosha!

I'm excited this app could also be embedded in remote stations in the forest (even as part of a trap) to collect data continuously. The possibilities for biodiversity survey are pretty exciting. 

Best of luck!

Users tagged:

Reply 2

Nirosha Priyadarshani Sep 30, 2015

Thanks John, of course our idea will provide solutions to a range of key problems in conservation, not just bird population monitoring. It is clear that the survival a bird population depends on reproduction, so enough balance between male and female is important. This idea can be used to classify (for example kiwi) calls to male and female. Or most probably even to identyfy individual birds...

So you are interested in predator controlling, and yes we can easily record them using programable recorders (like the ones DoC is currently using for kiwi surveys) and then use a bit advanced varsion of this app to monitor their abandance, pretty much in the same way as we process birdsongs. I can defenetely tell you that once we done this app, we are almost there. Meantime, I guess you can help us with recordings of these predators. During the development and testing we will need recordings of the target animals.

Indeed we are also excited to see the results!

Reply 1

Bruce McKinlay Oct 1, 2015

Have you had a look at similar work being done by Cornell Lab of Ornithology.  They have what looks like a similar tool in development.

Also have you thought  about how to integrate with NZ Birds online?  this online encyclopedia has lots of up to date bird information.  

Reply 0

Nirosha Priyadarshani Oct 2, 2015

Hi Bruce,

Thanks for your comments. 

NZ Birds online is one of my favourite libraries! It provides us with sample images and birdsongs. Also one of my supervisors, Dr Isabel Castro is a contributor to the NZ Birds online. We are in touch with some resource persons there like Les McPherson (McPherson Natural History Unit Sound Archive, www.archivebirdsnz.com) and really appreciate their support during last two years. As NZ Birds provides song examples from a wide range of NZ birds there is enough room to integrate our app there… There are also lots of other resources on the internet, and I also really like xeno-canto, which has amazing resources for worldwide bird calls. 

I think you are talking about Merlin Bird ID developed by Cornell Lab of Ornithology. Yes it is a great piece of work, but it is based on image processing, so you have to photograph the bird then ask the app to recognise it: they are interested in images but we are interested on their songs. They do mention some plans to have song recognition (and also host the Macaulay library, which is another amazing resource for birdsong recordings), but don’t seem to have done it yet, and even when they do it will be for common US birds.  

As I mentioned earlier in this post, developing this app is not the end of our journey, but is one step towards developing a sophisticated tool for bird population estimation and much more.

Reply 1

Share