Western Mediterranean Wetlands Bird Dataset

Manually labelled dataset of bird recordings from the species of interest inhabiting in the wetlands of the "Aiguamolls del Empordà" natural park in Girona, Spain. Concretely, the bird species that compose the dataset are:

Acrocephalus arundinaceus

Acrocephalus melanopogon

Acrocephalus scirpaceus

Alcedo atthis

Anas strepera

Anas platyrhynchos

Ardea purpurea

Botaurus stellaris

Charadrius alexandrinus

Ciconia ciconia

Circus aeruginosus

Coracias garrulus

Dendrocopos minor

Fulica atra

Gallinula chloropus

Himantopus himantopus

Ixobrychus minutus

Motacilla flava

Porphyrio porphyrio

Tachybaptus ruficollis

The dataset includes 5,795 annotated audio clips generated from a source of 1,098 recordings retrieved from the Xeno-Canto portal, adding up to a total of 201.6 minutes (12,096 seconds) of vocalizations of different lengths, alongside with their corresponding annotations.

And second, we also share the Mel spectrogram version of the dataset, where each image corresponds to 1-second window of the original audio, resulting in a total of 17,536 spectrogram images stored in matrix form in .npy files.

We would like to acknowledge all the Xeno-Canto community and their contributors for making the creation of the dataset possible. Specially, we would like to thank the following contributors for giving us special permission to use their recordings in this work despite having uploaded them on the Xeno-Cano portal under the terms BY-NC-ND: Anhäuser, Arnold Meijer, Bodo Sonnenburg, Chie-Jen Jerome Ko, Ding Li Yong, Eveny Luis, Fernand Deroussen (Sonothèque du MNHN), Hans Matheve, Herman van der Meer, Itziar Gutiérrez, Jacques Prevost, Jarek Matusiak, Jérémy Simar, Joost van Bruggen, Krzysztof Deoniziak, Lars Lachmann, Mandar Bhagat, Marc Anderson, Marco Dragonetti (www.birdsongs.it), Matthias, Feuersenger, Maudoc, Niels Krabbe, Patrick Franke, Peter Boesman, Piotr Szczypinski, Ruud van Beusekom.