MMEDIT: A Unified Approach to Audio Editing Using Audio Language Models

The paper introduces MMEDIT, a novel framework leveraging audio language models for versatile audio editing tasks. This research advances audio processing by providing a unified approach potentially simplifying complex editing workflows.

• •MMEDIT proposes a unified framework for multi-type audio editing.
• •It utilizes an audio language model for its core functionality.
• •The research originates from the ArXiv preprint server.