Birdsong recognition app launches using QMUL research
QMUL machine learning researcher Dan Stowell and his business partner Florence Wilkinson have launched Warblr, their mobile app that can automatically recognise birds by their song.
13 August 2015
Dan and Florence announced late last year that they had developed the machine learning technology to identify birds by their tweets. They launched a Kickstarter to further develop their prototype and received huge attention and celebrity support but feel short of their fundraising target.
QMUL’s Innovation Fund, which gives out funds from from the Engineering and Physical Sciences Research Council, stepped in to provide the necessary funds, and allowed the Warblr team to complete work on the iPhone app. The app is available to buy from iTunes now with the proceeds helping to fulfill the team’s ambitions to build an Android version, and to take Warblr to regions outside of the UK.
It is hoped that Warblr will help to increase interest in the natural world amongst young people in particular. The team also plan to make the data collected publically available for researchers and conservationists, to assist with monitoring and protecting our feathered friends.
Warblr co-founder Dan Stowell from QMUL’s Centre for Digital Music, says, "Here at Queen Mary University of London we’ve been at the forefront of developing this new technology for a while now, but as a scientist you don’t always get the chance to place your work directly in the hands of the general public – it’s a real privilege.
“And the beauty of it is that the more people who use Warblr and submit recordings to our server, the more accurate we can train our technology to become. In fact the benefits are wider than that – our users are citizen scientists, collecting data that will be shared with researchers and conservation organisations worldwide."
As well as audio intelligence, Warblr now features locations intelligence enabled by the British Trust for Ornithology with historic mapping data from the organisation, helping to refine the list of birds you’re likely to encounter through the app.
Warblr co-founder Florence Wilkinson says, “Warblr is the result of years of hard work, and we are so excited to finally be able to share it with our nation of bird-lovers! Whilst we didn’t make our target on Kickstarter, the community we built during the process have been incredible, forming a crack team of over 100 beta testers to help us refine the app, and even helping to write descriptions for the over 220 British bird species that Warblr can now recognise.
- Find out more about the app and download it via the Warblr website.
- Learn more about Dan's work in the School of Electronic Engineering and Computer Science.
For media information, contact:Neha Okhandiar
Public Relations Manager
Queen Mary University of London