English

Everybody Compose: Deep Beats To Music

Sound 2023-07-11 v1 Machine Learning Multimedia Audio and Speech Processing

Abstract

This project presents a deep learning approach to generate monophonic melodies based on input beats, allowing even amateurs to create their own music compositions. Three effective methods - LSTM with Full Attention, LSTM with Local Attention, and Transformer with Relative Position Representation - are proposed for this novel task, providing great variation, harmony, and structure in the generated music. This project allows anyone to compose their own music by tapping their keyboards or ``recoloring'' beat sequences from existing works.

Keywords

Cite

@article{arxiv.2306.06284,
  title  = {Everybody Compose: Deep Beats To Music},
  author = {Conghao Shen and Violet Z. Yao and Yixin Liu},
  journal= {arXiv preprint arXiv:2306.06284},
  year   = {2023}
}

Comments

Accepted MMSys '23

R2 v1 2026-06-28T11:01:41.094Z