10 datasets found

A dataset of 18,451 Chinese pop songs with precise syllable-note alignment.

A large-scale lyrics-melody dataset with 18,451 Chinese pop songs and 644,472 lyrics-context-melody triples.

Audeo: A full pipeline system for generating music from silent piano performance videos.

A music generation model that leverages free-form text as a conditioning factor, utilizing the diffusion model to generate waveform-based music.

The dataset consists of 4,235 lead sheets from the Wikifonia database containing melodies from genres including (but not limited to) jazz, folk, pop, and classical, and 120...

The proposed method is the first attempt that uses vector quantized diffusion models to generate symbolic music while providing users with flexible controls to generate symbolic...

The POP909 dataset is a pop song dataset of 909 MIDI pieces containing lead melodies, secondary melodies, piano accompaniment tracks, key signatures, and chord annotations.

The Lakh pianoroll dataset contains 6, 000 training and 973 evaluation piano sequences, with each music sequence spanning a length of 256 in total.

The JSB Chorale dataset is used for music generation learning from a single music segment.

The Maestro dataset contains MIDI and audio files of recorded performances from piano performance competitions.