Learning Jazz Pianist Style with Cross-Attention Conditioning
Researchers have developed a model that uses cross-attention conditioning to allow an AI to mimic the specific playing styles of twelve famous jazz pianists. By fine-tuning a transformer model on the PiJAMA dataset, the system can generate music that is accurately attributed to specific artists.
Why it matters
This project demonstrates advancements in style transfer and generative music, pushing the boundaries of how AI can interpret and replicate human artistic nuance.
Drew Edwards Akira Maezawa Simon Dixon
In 1994 Dick Hyman published In the Styles of... The Great Jazz Pianists : fifteen original tudes, each written in the manner of one master, from Scott Joplin to Bill Evans. Rather than transcribing their solos, Hyman composed new music that carries their signatures - Tatum's "rapid runs in both hands," Garner's "strumming, guitar-like left hand," Peterson's "tremolos and glissandi." That book is the inspiration for this project. Can a model learn to do what Hyman did: not just recognize who is playing, but play in their manner? Tatum, Garner, and Peterson are among the twelve pianists we study - and so is Hyman himself.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in