We have released the full test set & labels of FSDKaggle2019. This dataset was used for DCASE 2019 Task 2 - Audio tagging with noisy labels and minimal supervision, which was hosted on the Kaggle platform as a competition titled Freesound Audio Tagging 2019.
Both competition and dataset have been a collaboration between the Music Technology Group of Universitat Pompeu Fabra, and the Sound Understanding team at Google AI Perception. This effort was kindly sponsored by a Google Faculty Research Award 2018.
FSDKaggle2019 includes almost 30k clips encompassing 80 classes of the AudioSet Ontology, and allows development and evaluation of machine listening methods in conditions of label noise, minimal supervision, and real-world acoustic mismatch.
You can find more details in the Zenodo page and in our DCASE 2019 paper.
Download FSDKaggle2019 from Zenodo: https://doi.org/10.5281/zenodo.3612637
DCASE 2019 paper: E. Fonseca, M. Plakal, F. Font, D. P. W. Ellis, and X. Serra. Audio tagging with noisy labels and minimal supervision. Detection and Classification of Acoustic Scenes and Events (DCASE) Workshop, NYC, USA, 2019