MUSE：面向基于U-Net的语音增强的灵活声纹感受野与多路径融合增强型泰勒Transformer
MUSE: Flexible Voiceprint Receptive Fields and Multi-Path Fusion Enhanced Taylor Transformer for U-Net-based Speech Enhancement
AI总结：
针对语音增强中轻量化与高性能难平衡的问题，提出基于U-net的MUSE网络，通过带可变形嵌入的MET Transformer模块融合通道空间注意力分支，在VoiceBank+DEMAND数据集上以0.51M参数量实现有竞争力的性能并降低成本。
摘要：
在语音增强领域，实现轻量化设计与高性能之间的平衡仍是一项具有挑战性的任务。本文提出了基于U型网络（U-net）的多路径增强泰勒（MET）Transformer语音增强网络MUSE，这是一种构建于Unet架构之上的轻量化语音增强网络。该方法引入了新颖的多路径增强泰勒（MET）Transformer模块，模块集成可变形嵌入（DE），可为声纹提供灵活的感受野。MET Transformer采用独特设计，融合通道与空间注意力（CSA）分支，可促进通道信息交互，弥补泰勒Transformer框架内的空间注意力不足。通过在VoiceBank+DEMAND数据集上开展的大量实验，我们证明MUSE取得了具有竞争力的性能，同时大幅降低了训练与部署成本，其参数量仅为0.51M。
英文摘要：
Achieving a balance between lightweight design and high performance remains a challenging task for speech enhancement. In this paper, we introduce Multi-path Enhanced Taylor (MET) Transformer based U-net for Speech Enhancement (MUSE), a lightweight speech enhancement network built upon the Unet architecture. Our approach incorporates a novel Multi-path Enhanced Taylor (MET) Transformer block, which integrates Deformable Embedding (DE) to enable flexible receptive fields for voiceprints. The MET Transformer is uniquely designed to fuse Channel and Spatial Attention (CSA) branches, facilitating channel information exchange and addressing spatial attention deficits within the Taylor-Transformer framework. Through extensive experiments conducted on the VoiceBank+DEMAND dataset, we demonstrate that MUSE achieves competitive performance while significantly reducing both training and deployment costs, boasting a mere 0.51M parameters.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
