2步生成比50步更物理合理，靠锁定运动相位提升一致性
Physics in 2-Steps: Locking Motion Priors Before Visual Refinement Erases Them
- 用2步推断提取运动先验，全程锁定相位防止退化
- 物理一致性平均提升6.2分，视觉质量几乎不变
- 无需训练，兼容主流模型，大幅降低外部引导开销
图像到视频的扩散模型虽能生成高质量内容，但常违背物理规律。我们发现：相同模型下，2步生成的运动比50步输出更具物理一致性。通过谱分析发现，这是因为在去噪过程中相位显著退化（从第2步到第50步下降约18%），而幅度变化较小。基于此，我们提出PhaseLock——一种无需训练的框架，将仅2步推断中保留的有效运动先验，通过潜在增量引导（Latent Delta Guidance）贯穿整个去噪过程。该方法有效缓解相位退化，在多种模型上平均提升物理一致性6.2分，同时保持良好视觉质量，计算开销仅增加1.06倍时间、1.02倍内存，并减少约5倍对昂贵外部引导的依赖。
原文摘要 · Abstract (English)
Image-to-Video diffusion models leverage input images to generate visually stunning content, yet frequently produce motion that violates physical laws. We reveal a surprising finding: a 2-step generation often exhibits better physical consistency than a 50-step output from the same model. Through spectral analysis, we trace this to phase erosion during denoising; the phase degrades significantly (dropping by $\approx 18\%$ from step 2 to step 50), whereas the magnitude remains relatively stable. Building on this insight, we propose PhaseLock, a training-free framework that preserves the valid motion priors from few-step inference throughout the denoising trajectory. Rather than relying on full-step inference for physical consistency, PhaseLock extracts a motion prior from just 2 steps and enforces it onto high-fidelity generation via Latent Delta Guidance. Our approach effectively mitigates phase degradation, improving physical consistency by an average of 6.2 points across diverse models while largely maintaining visual fidelity, with negligible overhead ($1.06\times$ time, $1.02\times$ memory) and reduced reliance on expensive external guidance methods ($\sim5\times$ time). Project Page: https://dnwjddl.github.io/phaselock
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。