TL;DR在这篇文章中，我们发现了扩散模型生成质量受到迭代次数限制的根本原因，并提出了一个简单而有效的解决方案来缓解这些影响。我们的解决方案可以应用于任何现有的扩散模型，并且在各种 SOTA 体系结构上运行多个数据集和配置进行实验和详尽的消融研究，证明能够立即提高它们的生成质量。
Abstract
iterative denoising-based generation, also known as denoising diffusion
models, has recently been shown to be comparable in quality to other classes of
generative models, and even surpass them. Including, in particular, Generative
Adversarial Networks, which are currently the state of