Identification of bird species from their singing

Biodiversity is one of the mysterious phenomenon that we are understanding more and more as time passes by. Aiding the process of understanding nature is an initiative by BIOTOPE society; an open platform to bring together researchers to retrieve information by identifying bird species from calls or singing of different bird species. The main idea behind this work is to identify bird species from their calls or songs. The work presented in this article explores the automated method to identify bird species based on songs and calls of different bird species. This will help in archiving and identifying new species or subspecies. Moreover, it will help behavioral and ecological studies of birds and aid to the surveillance of bird population. In this work we implement a system that will be able to automatically identify bird species from acoustic recordings. Proposed system's ROC performance analysis shows that using Random Decision Tree we can yield 0.96 area under the curve.