The Piano Inpainting Application

  • 2021-07-13 09:33:11
  • Gaëtan Hadjeres, Léopold Crestel
  • 46

Abstract

Autoregressive models are now capable of generating high-quality minute-longexpressive MIDI piano performances. Even though this progress suggests newtools to assist music composition, we observe that generative algorithms arestill not widely used by artists due to the limited control they offer,prohibitive inference times or the lack of integration within musicians'workflows. In this work, we present the Piano Inpainting Application (PIA), agenerative model focused on inpainting piano performances, as we believe thatthis elementary operation (restoring missing parts of a piano performance)encourages human-machine interaction and opens up new ways to approach musiccomposition. Our approach relies on an encoder-decoder Linear Transformerarchitecture trained on a novel representation for MIDI piano performancestermed Structured MIDI Encoding. By uncovering an interesting synergy betweenLinear Transformers and our inpainting task, we are able to efficiently inpaintcontiguous regions of a piano performance, which makes our model suitable forinteractive and responsive A.I.-assisted composition. Finally, we introduce ourfreely-available Ableton Live PIA plugin, which allows musicians to smoothlygenerate or modify any MIDI clip using PIA within a widely-used professionalDigital Audio Workstation.

 

Quick Read (beta)

loading the full paper ...