arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2406.00294cs.SDcs.LGeess.AS
收藏

通过合成器编程实现创意文本到音频的生成

Creative Text-to-Audio Generation via Synthesizer Programming

Massachusetts Institute of Technology(麻省理工学院);Media Lab(媒体实验室)

Manuel Cherep, Nikhil Singh, Jessica Shand

AI总结:

提出CTAG方法,利用仅78个参数的虚拟模块化合成器迭代更新参数以生成文本提示的高质量音频,解决了传统神经音频合成难以微调的问题,生成的声音独特且具艺术性,可识别性与现有模型相当。

摘要:

神经音频合成方法现在允许使用自然语言来指定想法。然而,这些方法产生的结果难以轻易微调,因为它们基于大型潜在空间和多达数十亿个不可解释的参数。我们提出了一种文本到音频生成方法,该方法利用仅包含 78 个参数的虚拟模块化声音合成器。由于其灵活性和直观的控制,合成器长期以来一直被熟练的声音设计师用于音乐和电影等媒体。我们的方法 CTAG 迭代地更新合成器的参数,以生成文本提示的高质量音频渲染,这些音频可以轻松检查和微调。以这种方式产生的声音也更抽象,捕捉基本的概念特征而非细粒度的声学细节,类似于简单的草图如何生动地传达视觉概念。我们的结果表明,CTAG 产生的声音独特、被视为具有艺术性,且与最近的神经音频合成模型具有相似的可识别性,这使其成为一款有价值的互补工具。

英文摘要:

Neural audio synthesis methods now allow specifying ideas in natural language. However, these methods produce results that cannot be easily tweaked, as they are based on large latent spaces and up to billions of uninterpretable parameters. We propose a text-to-audio generation method that leverages a virtual modular sound synthesizer with only 78 parameters. Synthesizers have long been used by skilled sound designers for media like music and film due to their flexibility and intuitive controls. Our method, CTAG, iteratively updates a synthesizer's parameters to produce high-quality audio renderings of text prompts that can be easily inspected and tweaked. Sounds produced this way are also more abstract, capturing essential conceptual features over fine-grained acoustic details, akin to how simple sketches can vividly convey visual concepts. Our results show how CTAG produces sounds that are distinctive, perceived as artistic, and yet similarly identifiable to recent neural audio synthesis models, positioning it as a valuable and complementary tool.

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。

↑