All across the globe, countless species are becoming endangered or even extinct. In order to halt the loss of species, we must understand the underlying reasons, and keep track of biodiversity over time and in different habitats. Birds are often considered important indicators of habitat changes. They also have a great advantage: most of them produce characteristic sounds within the range of human hearing – which means they can be detected even if they are not to be seen.
During the famous dawn chorus, many birds are highly vocally active, which means that different individuals of the same and of different species sing at the same time. This makes the analysis of the collected dawn chorus data an enormous challenge. Our aim is to collaborate with bird song experts, IT and data scientists to eventually build artificial intelligence for automatic species recognition.
We are currently in the stage of analyzing some of the recordings by human hand (or better by ear), to build a library which the computer algorithms will be able to learn from. This means, we do not yet have any results on species distributions to show you yet. However, we can show you the species that our citizen scientists reported (in German only):