Skip to content

Dataset · Fhrozen

Freesound labeled sound events (50K)

Fhrozen/FSD50k

Mirror of the well-known FSD50K dataset for recognizing and classifying everyday and natural sounds.

What's inside

Inside are audio recordings with manually labeled sound events, totaling between 10 and 100 thousand examples. This is a copy of the original open dataset from the Zenodo platform, created by researchers based on Freesound.

What you can do with it

  • Training sound classification and acoustic event models
  • Creating noise and acoustic environment monitoring systems
Tasks
Audio classification
Size
10K–100K rows
Downloads this month
66.8K
Likes
32

How to load it

from datasets import load_dataset

ds = load_dataset("Fhrozen/FSD50k")
Freesound labeled sound events (50K): dataset Fhrozen/FSD50k | NeyroLearn