AudioEditor: A Training-Free Diffusion-Based Audio Editing Framework

TL;DR

AudioEditor is a novel training-free framework for precise and high-quality audio editing using pretrained diffusion models, addressing challenges of preserving unedited audio parts while executing accurate edits.

Contribution

It introduces a training-free audio editing method leveraging Null-text Inversion and EOT-suppression techniques on pretrained diffusion models, a novel approach in audio editing.

Findings

Effective preservation of original audio features during editing

High-quality audio edits validated through objective and subjective tests

Demonstrates the feasibility of training-free audio editing with diffusion models

Abstract

Diffusion-based text-to-audio (TTA) generation has made substantial progress, leveraging latent diffusion model (LDM) to produce high-quality, diverse and instruction-relevant audios. However, beyond generation, the task of audio editing remains equally important but has received comparatively little attention. Audio editing tasks face two primary challenges: executing precise edits and preserving the unedited sections. While workflows based on LDMs have effectively addressed these challenges in the field of image processing, similar approaches have been scarcely applied to audio editing. In this paper, we introduce AudioEditor, a training-free audio editing framework built on the pretrained diffusion-based TTA model. AudioEditor incorporates Null-text Inversion and EOT-suppression methods, enabling the model to preserve original audio features while executing accurate edits. Comprehensive objective and subjective experiments validate the effectiveness of AudioEditor in delivering high-quality audio edits. Code and demo can be found at https://github.com/NKU-HLT/AudioEditor.