Speech-Copilot: Leveraging Large Language Models for Speech Processing via Task Decomposition, Modularization, and Program Generation

TL;DR

Speech-Copilot introduces a modular, instruction-driven framework for speech processing that leverages large language models to decompose tasks, build toolsets, and generate programs, achieving state-of-the-art results without additional training.

Contribution

It presents a novel modular framework that constructs speech processing toolsets from instructions and uses large language models for task execution, bypassing end-to-end training.

Findings

Achieved state-of-the-art performance on Dynamic-SUPERB benchmark

Effectively decomposes complex speech tasks into manageable sub-tasks

Provides a flexible, extendable approach without additional training

Abstract

In this work, we introduce Speech-Copilot, a modular framework for instruction-oriented speech-processing tasks that minimizes human effort in toolset construction. Unlike end-to-end methods using large audio-language models, Speech-Copilot builds speech processing-specific toolsets by analyzing pre-collected task instructions and breaking tasks into manageable sub-tasks. It features a flexible agent based on large language models that performs tasks through program generation. Our approach achieves state-of-the-art performance on the Dynamic-SUPERB benchmark, demonstrating its effectiveness across diverse speech-processing tasks. Key contributions include: 1) developing an innovative framework for speech processing-specific toolset construction, 2) establishing a high-performing agent based on large language models, and 3) offering a new perspective on addressing challenging instruction-oriented speech-processing tasks. Without additional training processes required by end-to-end approaches, our method provides a flexible and extendable solution for a wide range of speech-processing applications.