What’s Out There
← Back to swiping
Audio & musicFreeAcademic

FSD50K

Fonseca, Favory, Pons, et al. (Universitat Pompeu Fabra)

51,197 Freesound clips labeled with 200 sound-event classes drawn from the AudioSet Ontology.

An open set of human-labeled sound events: 51,197 Freesound audio clips spread unevenly over 200 classes from the AudioSet Ontology, totaling about 108 hours. Labels were gathered with crowd annotation and validated. It is used to train and benchmark audio tagging and sound event detection.

Access

Cost
FreeFree download from Zenodo; no account needed.
License
Provider's own termsPer-clip Creative Commons licenses (CC0, CC BY, CC BY-NC, and CC Sampling+) plus dataset terms
Sign-up
No sign-up
How you get it
Bulk downloadWAV, ZIP (split archive)

What's in it

One row is
One audio clip with one or more sound-event labels
Covers
Not geographic
Years
Clips from Freesound (release 2020)
Updated
One-time release
Size
51,197 clips in 200 classes
sound classaudio clipclip durationoriginal Freesound IDlicense of clip

Quality

4 of 5

Well documented with a published paper and standard train and test splits; static since 2020.

  • Mixed clip licenses; check each clip's terms for commercial use
  • Strong class imbalance

Connects to

  • AudioSet: Larger weakly labeled YouTube set that shares the ontology
  • Freesound: Source repository of the audio clips
  • FSDKaggle2018: Earlier smaller subset from the same project

Questions it could answer

  • Which sound classes are hardest to tag?
  • How does a model pretrained on AudioSet do on FSD50K labels?

Similar data sets

  • AudioSet

    Two million 10-second YouTube clips labeled with 527 kinds of sound, from barking to bagpipes.

    Audio & musicGoogle Research (Sound Understanding group)Free

  • MusicNet

    330 classical recordings with over a million labeled notes aligned to the second.

    Audio & musicThickstun, Harchaoui, Kakade (University of Washington)Free

  • GuitarSet

    360 guitar excerpts with hexaphonic pickups, giving note-level labels for each string.

    Audio & musicXi, Bittner, Pauwels, Ye, Bello (NYU MARL)Free

  • LibriSpeech ASR Corpus

    1,000 hours of volunteers reading public-domain audiobooks aloud, the speech recognition yardstick.

    Audio & musicOpenSLR (Vassil Panayotov, Daniel Povey, and colleagues)Free

  • MUSDB18: a corpus for music separation

    150 full songs delivered as a mix plus separated vocals, drums, bass, and other stems.

    Audio & musicRafii, Liutkus, Stöter, et al. (SigSep)Free

  • Mozilla Common Voice

    Volunteers reading sentences aloud in 100+ languages, donated to the public domain.

    Audio & musicMozilla FoundationFree