Text Prompt is Not Enough: Sound Event Enhanced Prompt Adapter for Target Style Audio Generation

TL;DR

This paper introduces a novel sound event enhanced prompt adapter for multi-style audio generation that combines text and audio references, achieving state-of-the-art results and better style control.

Contribution

It proposes a new adaptive style transfer method using cross-attention and layer normalization, along with a new dataset for dual-prompt audio generation.

Findings

Achieved state-of-the-art Fréchet Distance of 26.94

Attained KL Divergence of 1.82

Generated audio closely matches reference styles

Abstract

Current mainstream audio generation methods primarily rely on simple text prompts, often failing to capture the nuanced details necessary for multi-style audio generation. To address this limitation, the Sound Event Enhanced Prompt Adapter is proposed. Unlike traditional static global style transfer, this method extracts style embedding through cross-attention between text and reference audio for adaptive style control. Adaptive layer normalization is then utilized to enhance the model's capacity to express multiple styles. Additionally, the Sound Event Reference Style Transfer Dataset (SERST) is introduced for the proposed target style audio generation task, enabling dual-prompt audio generation using both text and audio references. Experimental results demonstrate the robustness of the model, achieving state-of-the-art Fr\'echet Distance of 26.94 and KL Divergence of 1.82, surpassing Tango, AudioLDM, and AudioGen. Furthermore, the generated audio shows high similarity to its corresponding audio reference. The demo, code, and dataset are publicly available.