SpeechCraft: A Fine-grained Expressive Speech Dataset with Natural Language Description

TL;DR

SpeechCraft is a large-scale, fine-grained expressive speech dataset with natural language descriptions, created using an automatic annotation system, to enhance speech style understanding and synthesis tasks.

Contribution

The paper introduces an automatic annotation system that generates detailed natural language descriptions for speech clips, enabling the creation of a large, diverse expressive speech dataset.

Findings

The dataset contains approximately 2,000 hours of speech data.

SpeechCraft improves performance in speech style understanding tasks.

It significantly boosts stylistic speech synthesis capabilities.

Abstract

Speech-language multi-modal learning presents a significant challenge due to the fine nuanced information inherent in speech styles. Therefore, a large-scale dataset providing elaborate comprehension of speech style is urgently needed to facilitate insightful interplay between speech audio and natural language. However, constructing such datasets presents a major trade-off between large-scale data collection and high-quality annotation. To tackle this challenge, we propose an automatic speech annotation system for expressiveness interpretation that annotates in-the-wild speech clips with expressive and vivid human language descriptions. Initially, speech audios are processed by a series of expert classifiers and captioning models to capture diverse speech characteristics, followed by a fine-tuned LLaMA for customized annotation generation. Unlike previous tag/templet-based annotation frameworks with limited information and diversity, our system provides in-depth understandings of speech style through tailored natural language descriptions, thereby enabling accurate and voluminous data generation for large model training. With this system, we create SpeechCraft, a fine-grained bilingual expressive speech dataset. It is distinguished by highly descriptive natural language style prompts, containing approximately 2,000 hours of audio data and encompassing over two million speech clips. Extensive experiments demonstrate that the proposed dataset significantly boosts speech-language task performance in stylist speech synthesis and speech style understanding.