Free Music Archive Audio Tracks Across Eight Genres

FMA Genre Classification Dataset contains 8,000 audio tracks from the Free Music Archive, each 30 seconds long. The tracks are evenly distributed across 8 genres, curated by author rpmon and last updated in December 2024.

Use Cases

• Train a genre classifier using the 8 genre labels on 8,000 audio tracks.
• Analyze audio feature distributions across the 8 distinct music genres.
• Benchmark audio classification models on a balanced dataset of 8,000 30-second samples.

Strengths

• Contains 8,000 audio tracks, providing a substantial sample for model training.
• Tracks are evenly distributed across 8 genres, ensuring a balanced class structure.
• Each audio sample is a consistent 30 seconds in length, simplifying preprocessing.

Limitations

• The dataset is a 'small' subset of the larger Free Music Archive, limiting its scope.
• Specific audio features or metadata columns are not detailed in the provided input.
• The temporal coverage and geographic origin of the music tracks are unknown.

Provenance

Source

Free Music Archive (FMA), subset curated by rpmon.

Collection Method

Subset extracted from the Free Music Archive; specific curation method unknown.

Freshness

Last updated on 2024-12-14.