GuitarFlow：基于流匹配与风格迁移的吉他谱谱实用电吉他合成
声音
2025-10-28 v1 人工智能
音频与语音处理
摘要
人工智能在音频领域的音乐生成近年来取得了稳健的进展。然而对于某些乐器，特别是吉他，可控乐器合成在表现力上仍受限。我们提出 GuitarFlow，一个专为电吉他合成设计的模型。生成过程通过谱谱（tablature）进行引导，这是一种常见且直观的吉他特定符号格式。谱谱格式能够轻松地表示吉他特定的演奏技巧（如滑音、静音弦和连奏），而其他常见的音乐记谱格式如 MIDI 则难以表示。本模型依赖于一个中介步骤：首先使用简单的基于样本的虚拟乐器将谱谱渲染为音频，然后通过流匹配进行风格迁移，将虚拟乐器音频转换为更真实的示例。 resulting in a model that is quick to train and to perform inference, requiring less than 6 hours of training data. We present the results of objective evaluation metrics, together with a listening test, in which we show significant improvement in the realism of the generated guitar audio from tablatures.
关键词
引用
@article{arxiv.2510.21872,
title = {GuitarFlow: Realistic Electric Guitar Synthesis From Tablatures via Flow Matching and Style Transfer},
author = {Jackson Loth and Pedro Sarmento and Mark Sandler and Mathieu Barthet},
journal= {arXiv preprint arXiv:2510.21872},
year = {2025}
}
备注
To be published in Proceedings of the 17th International Symposium on Computer Music and Multidisciplinary Research (CMMR)