提出时间对齐引导机制，让扩散模型生成更精准的样本。
Temporal Alignment Guidance: On-Manifold Sampling in Diffusion Models
- 用时间预测器实时检测生成偏差，动态修正轨迹。
- 在各时间步均提升样本与数据流形对齐度，改善生成质量。
- 适合需要高保真生成的图像/视频任务，尤其对抗性引导场景。
扩散模型虽在生成任务中表现卓越，但即使训练良好，生成过程仍会累积误差。当引入任意引导以控制生成属性时，常导致样本失真。本文提出一种通用解决方案，应对扩散模型中的离流形现象。通过时间预测器估算每一步与目标数据流形的偏离，发现时间差距越大，生成质量越低。据此设计新型引导机制——时间对齐引导（TAG），在生成过程中每一步都将样本拉回目标流形。大量实验表明，TAG在各时间步均使样本紧密贴合目标流形，在多个下游任务中显著提升生成质量。
原文摘要 · Abstract (English)
Diffusion models have achieved remarkable success as generative models. However, even a well-trained model can accumulate errors throughout the generation process. These errors become particularly problematic when arbitrary guidance is applied to steer samples toward desired properties, which often breaks sample fidelity. In this paper, we propose a general solution to address the off-manifold phenomenon observed in diffusion models. Our approach leverages a time predictor to estimate deviations from the desired data manifold at each timestep, identifying that a larger time gap is associated with reduced generation quality. We then design a novel guidance mechanism, `Temporal Alignment Guidance' (TAG), attracting the samples back to the desired manifold at every timestep during generation. Through extensive experiments, we demonstrate that TAG consistently produces samples closely aligned with the desired manifold at each timestep, leading to significant improvements in generation quality across various downstream tasks.
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。