Musical Speech: A Transformer-based Composition Tool

TL;DR

This paper introduces a Transformer-based tool that converts user-provided speech into musical outlines, enabling personalized music creation with a clear link between speech and music, without needing paired datasets.

Contribution

The paper presents a novel pipeline combining speech processing, heuristics, and Transformer models for music generation from speech without requiring paired training data.

Findings

Effective music generation from speech demonstrated

Tool allows user customization with speech input

No paired dataset needed for training

Abstract

In this paper, we propose a new compositional tool that will generate a musical outline of speech recorded/provided by the user for use as a musical building block in their compositions. The tool allows any user to use their own speech to generate musical material, while still being able to hear the direct connection between their recorded speech and the resulting music. The tool is built on our proposed pipeline. This pipeline begins with speech-based signal processing, after which some simple musical heuristics are applied, and finally these pre-processed signals are passed through Transformer models trained on new musical tasks. We illustrate the effectiveness of our pipeline -- which does not require a paired dataset for training -- through examples of music created by musicians making use of our tool.