Los Angeles MIDI Dataset for Music AI Research

Los Angeles MIDI Dataset is a collection of MIDI files for music information retrieval and AI purposes, described as a state-of-the-art kilo-scale resource. It was created by projectlosangeles and was last updated in February 2024.

Use Cases

• Train models for symbolic music generation using MIDI note sequences and instrument data.
• Perform music information retrieval tasks like genre classification using metadata tags.
• Develop and benchmark models for audio synthesis from symbolic MIDI representations.
• Filter and search MIDI files by attributes like tempo, key, or instrumentation for dataset curation.

Strengths

• Described as a 'kilo-scale' dataset, implying it contains thousands of MIDI files.
• Promoted as a state-of-the-art (SOTA) resource for its intended domain.

Limitations

• Exact size, row count, and specific column/feature details are not provided.
• The description is promotional; independent validation of quality and scale is unavailable.

Provenance

Source

projectlosangeles on Hugging Face.

Collection Method

Potentially created from a 'MIDI scrape' as mentioned in the description.

Freshness

Last updated on 2024-02-20.

Geography

The 'Los Angeles' moniker may not indicate geographic coverage; likely a project name.