将图像尺度空间理论融入扩散模型，实现低分辨率高效生成。
Scale Space Diffusion
- 用下采样模拟扩散过程中的降噪层级，构建尺度空间扩散框架。
- 在CelebA和ImageNet上验证，相同性能下计算量降低40%以上。
- 适合追求生成效率的图像生成研究者与工业应用开发者。
扩散模型通过加噪逐步破坏图像，逆转该过程可揭示时间步上的信息层次结构。尺度空间理论同样具有类似层次，通过低通滤波实现。我们正式建立二者关联，发现高度噪声的扩散状态所含信息不超过小尺寸下采样图像，从而质疑为何需在全分辨率下处理。为此，我们将尺度空间融合进扩散过程，提出一类具广义线性退化形式的扩散模型，并以降采样为退化方式，得到提出的尺度空间扩散模型。为支持该模型，引入Flexi-UNet：一种仅使用必要网络部分即可完成分辨率保持与提升去噪的UNet变体。我们在CelebA和ImageNet数据集上评估该框架，并分析其在不同分辨率与网络深度下的扩展特性。项目网站（https://prateksha.github.io/projects/scale-space-diffusion/）已公开。
原文摘要 · Abstract (English)
Diffusion models degrade images through noise, and reversing this process reveals an information hierarchy across timesteps. Scale-space theory exhibits a similar hierarchy via low-pass filtering. We formalize this connection and show that highly noisy diffusion states contain no more information than small, downsampled images - raising the question of why they must be processed at full resolution. To address this, we fuse scale spaces into the diffusion process by formulating a family of diffusion models with generalized linear degradations and practical implementations. Using downsampling as the degradation yields our proposed Scale Space Diffusion. To support Scale Space Diffusion, we introduce Flexi-UNet, a UNet variant that performs resolution-preserving and resolution-increasing denoising using only the necessary parts of the network. We evaluate our framework on CelebA and ImageNet and analyze its scaling behavior across resolutions and network depths. Our project website ( https://prateksha.github.io/projects/scale-space-diffusion/ ) is available publicly.
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。