无需训练即可生成无限长音频，解决拼接导致的不连贯问题。
InfiniteAudio: Infinite-Length Audio Generation with Consistency
- 采用先进先出推理策略，固定输入大小实现持续生成。
- 在多个指标上表现优于或相当现有方法，生成音频更连贯。
- 适合需要长时间高质量音频生成的研究与应用者。
本文提出InfiniteAudio，一种基于扩散模型的文本到音频方法，可生成无限长度音频。当前方法受限于内存，因输出随输入增长而增大，难以实现长时生成。常见方案是拼接短段音频，但缺乏共享时间上下文，易产生不一致。InfiniteAudio无缝集成于现有流程，无需额外训练。提出两项关键技术：FIFO采样（固定尺寸输入的先进先出推理策略）和曲线去噪（选择性优先关键去噪步骤以提升效率）。实验表明，InfiniteAudio在所有评估指标上均达到或超越现有方法。音频样本可在项目页面获取。
原文摘要 · Abstract (English)
This paper presents InfiniteAudio, a simple yet effective strategy for generating infinite-length audio using diffusion-based text-to-audio methods. Current approaches face memory constraints because the output size increases with input length, making long duration generation challenging. A common workaround is to concatenate short audio segments, but this often leads to inconsistencies due to the lack of shared temporal context. To address this, InfiniteAudio integrates seamlessly into existing pipelines without additional training. It introduces two key techniques: FIFO sampling, a first-in, first-out inference strategy with fixed-size inputs, and curved denoising, which selectively prioritizes key diffusion steps for efficiency. Experiments show that InfiniteAudio achieves comparable or superior performance across all metrics. Audio samples are available on our project page.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。