动态调整负向提示，让AI绘图更精准不跑偏
Dynamic Negative Guidance of Diffusion Models
- 根据生成过程状态实时调节负向引导强度
- 在MNIST和CIFAR10上提升图像安全性和质量
- 无需额外训练，可直接用于Stable Diffusion
负向提示（NP）在扩散模型中广泛用于避免生成不需要的特征，尤其在文本到图像应用中。本文指出，传统NP受限于恒定引导系数的假设，因反向过程具有非平稳性和状态依赖性，可能导致结果极差甚至失败。基于此，我们提出一种原则性方法——动态负向引导（DNG），通过时间与状态相关的近优调制实现引导，无需额外训练。与普通NP不同，DNG需在去噪过程中估计后验类别概率，通过追踪生成过程中的离散马尔可夫链，仅增加少量计算开销。我们在MNIST和CIFAR10上评估了DNG在类移除任务上的表现，结果显示其相比基线方法提升了安全性、类别平衡性和图像质量。此外，将DNG应用于Stable Diffusion可获得更准确且侵入性更低的引导效果。
原文摘要 · Abstract (English)
Negative Prompting (NP) is widely utilized in diffusion models, particularly in text-to-image applications, to prevent the generation of undesired features. In this paper, we show that conventional NP is limited by the assumption of a constant guidance scale, which may lead to highly suboptimal results, or even complete failure, due to the non-stationarity and state-dependence of the reverse process. Based on this analysis, we derive a principled technique called Dynamic Negative Guidance, which relies on a near-optimal time and state dependent modulation of the guidance without requiring additional training. Unlike NP, negative guidance requires estimating the posterior class probability during the denoising process, which is achieved with limited additional computational overhead by tracking the discrete Markov Chain during the generative process. We evaluate the performance of DNG class-removal on MNIST and CIFAR10, where we show that DNG leads to higher safety, preservation of class balance and image quality when compared with baseline methods. Furthermore, we show that it is possible to use DNG with Stable Diffusion to obtain more accurate and less invasive guidance than NP.
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。