为每张图定制生成路径，提速同时保质保控。
RayFlow: Instance-Aware Diffusion Acceleration via Adaptive Flow Trajectories
- 为每个样本设计专属生成轨迹，动态调整采样路径。
- 仅用20步即可生成高质量图像，速度显著提升。
- 适合需要快速生成且保持多样性的图像应用。
扩散模型在多个领域取得显著成果，但生成速度慢仍是关键挑战。现有加速方法虽减少采样步骤，却常牺牲样本质量、可控性或引入训练复杂度。为此，我们提出RayFlow，一种新型扩散框架，通过为每个样本引导至特定目标分布的独立路径，实现最小化采样步数的同时，保持生成多样性与稳定性。此外，我们引入Time Sampler，一种重要性采样技术，聚焦关键时间步以提升训练效率。大量实验表明，相较于现有加速技术，RayFlow在生成高质量图像方面具有更快的速度、更强的控制能力及更高的训练效率。
原文摘要 · Abstract (English)
Diffusion models have achieved remarkable success across various domains. However, their slow generation speed remains a critical challenge. Existing acceleration methods, while aiming to reduce steps, often compromise sample quality, controllability, or introduce training complexities. Therefore, we propose RayFlow, a novel diffusion framework that addresses these limitations. Unlike previous methods, RayFlow guides each sample along a unique path towards an instance-specific target distribution. This method minimizes sampling steps while preserving generation diversity and stability. Furthermore, we introduce Time Sampler, an importance sampling technique to enhance training efficiency by focusing on crucial timesteps. Extensive experiments demonstrate RayFlow's superiority in generating high-quality images with improved speed, control, and training efficiency compared to existing acceleration techniques.
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。