An Open Dataset for Research on Audio Field Recording Archives: freefield1010

By Dan Stowell, Mark D. Plumbley. 2013

68 citations recorded by the source corpus.

We introduce a free and open dataset of 7690 audio clips sampled from the field-recording tag in the Freesound audio archive. The dataset is designed for use in research related to data mining in audio archives of field recordings / soundscapes. Audio is standardised, and audio and metadata are Creative Commons licensed. We describe the data preparation process, characterise the dataset descriptively, and illustrate its use through an auto-tagging experiment.

Subfields: Computer Vision and Pattern Recognition, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Signal Processing

Citations in the corpus

FSD50K: An Open Dataset of Human-Labeled Sound Events (2020, methodology): “Several datasets containing Freesound audio have been widely used by the research community [22, 32, 33, 39, 57, 58], showing its usefulness for dataset creation.”
Freesound Datasets: A Platform for the Creation of Open Audio Datasets (2017, background): “A number of openly available datasets containing Freesound clips have already been used for research [24, 30, 34], showing that it is already a useful source for the creation of datasets.”
Detection and Classification of Acoustic Scenes and Events (2015, background): “1 However, these vary widely in recording conditions, recording quality and file format [6], [37], and so were unsuitable for our experimental aim to evaluate systems run with consistent audio front-end.”
Chime-home: A dataset for sound source recognition in a domestic environment (2015, methodology): “Notably, our dataset differs from [15, 16], being based on domestic environment recordings for the intended applications of human activity monitoring and voice detection; further our annotation protocol differs from [14] in that we obtain annotations with the aid of multiple annotators, among which…”
BirdNET: A deep learning solution for avian diversity monitoring (2021, methodology): “We used 16 classes from the AudioSet and enriched them with recordings from the Freefield1010 (Stowell and Plumbley, 2013) and WarblR datasets, which are both part of the DCASE bird detection challenge and did not contain bird sounds (Mesaros et al., 2018; Stowell et al., 2016).”
SONYC Urban Sound Tagging (SONYC-UST): A Multilabel Dataset from an Urban Acoustic Sensor Network (2019, background): “The freefield1010 [6], UrbanSound [7], UrbanSound8k, [7], and Urban-SED [8] datasets contain recordings curated from Freesound [9] rather than recorded in a realistic noise monitoring scenario.”