Denoising diffusion models (DDMs) have attracted attention for their exceptional generation quality and diversity. This success is largely attributed to the use of class- or text-conditional diffusion guidance methods, such as classifier and classifier-free guidance. In this paper, we present a more comprehensive perspective that goes beyond the traditional guidance methods. From this generalized perspective, we introduce novel condition- and training-free strategies to enhance the quality of generated images. As a simple solution, blur guidance improves the suitability of intermediate samples for their fine-scale information and structures, enabling diffusion models to generate higher quality samples with a moderate guidance scale. Improving upon this, Self-Attention Guidance (SAG) uses the intermediate self-attention maps of diffusion models to enhance their stability and efficacy. Specifically, SAG adversarially blurs only the regions that diffusion models attend to at each iteration and guides them accordingly. Our experimental results show that our SAG improves the performance of various diffusion models, including ADM, IDDPM, Stable Diffusion, and DiT. Moreover, combining SAG with conventional guidance methods leads to further improvement.

翻译：去噪扩散模型凭借其卓越的生成质量和多样性而备受关注。这一成功在很大程度上归因于类条件或文本条件的扩散引导方法（如分类器引导和无分类器引导）。本文提出了一种超越传统引导方法的更全面视角。基于这一广义视角，我们引入了无需条件和训练的新型策略，以增强生成图像的质量。作为一种简单方案，模糊引导通过改善中间样本在其精细尺度信息与结构上的适配性，使扩散模型能够以适中的引导尺度生成更高质量的样本。在此基础上，自注意力引导进一步利用扩散模型的中间自注意力图来提升其稳定性与有效性。具体而言，SAG在每次迭代中仅对扩散模型关注的区域施加对抗性模糊，并据此进行引导。实验结果表明，我们的SAG方法提升了包括ADM、IDDPM、Stable Diffusion和DiT在内的多种扩散模型的性能。此外，将SAG与传统引导方法结合可带来进一步的性能提升。