Datasets

Scaled singing dataset introduced with ACE-Opencpop for multi-singer voice synthesis research.

Scaled multi-singer Mandarin singing dataset introduced for large-scale singing-voice synthesis.

Crowdsourced low- and high-level acoustic descriptors indexed by MusicBrainz recording identifiers.

Four genre-annotation datasets joining AcousticBrainz features with AllMusic, Discogs, Last.fm and Tagtraum labels.

Russian-language directory and practical resource for AI music, audio, DAW, analysis, conversion, rhyme, and prompt tools.

Manual beat, melody and chorus annotations for RWC musical pieces.

EPFL DCML corpus of annotated Beethoven piano-sonata scores for computational musicology.

Crowdsourced natural-language captions paired with AudioSet-derived audio clips.

Human-labeled ten-second YouTube segments spanning music, speech and environmental sound classes.

Alliance-operated AI-music chart product; the accessible chart emphasized mainland-China tracks and exposes limited methodology.

Azerbaijani platform presenting registration, provenance, analysis, and protection services for AI-created or altered content including voice and music.

User-entered melodies and model-generated four-voice harmonizations contributed through the 2019 Bach Doodle.

UGent dataset pairing reference music with television query audio for broadcast fingerprinting research.

Annual artist survey from BAM! Popauteurs and academic collaborators covering income, working conditions, and generative-AI use and concern.

Conjunto de Dados de Séries Temporais de Músicas Brasileiras

Dataset combining Spotify-derived acoustic features and processed lyrics for approximately 51,000 Brazilian tracks.

Open MIR dataset of Brazilian rhythmic instruments, solo and multi-instrument recordings across five rhythm styles.