无需训练即可加速流模型，提升生成速度且保持画质。
VDE: Training-Free Accelerating Rectified Flow Model via Velocity Decomposition and Estimation
- 将速度分解为输入相关与无关分量，实现精准自适应估计。
- 在图像视频生成中提速3.22倍，LPIPS仅0.069，画质损失极小。
- 适合需要快速部署的生成任务，尤其适用于已训练好的流模型。
尽管修正流模型在图像、视频和3D生成中表现优异，但其实际应用受限于推理速度慢。现有加速方法依赖缓存历史特征，忽视静态缓存与动态输入间的偏差，导致输出质量下降。本文提出无需训练的加速方法VDE（Velocity Decomposition and Estimation），将模型速度分解为与输入平行和正交的分量，利用其时间可预测性和方向稳定性，实现精准、输入自适应的估计。为防止误差累积，定期通过完整前向传播锚定模型状态。在图像与视频生成任务上的大量实验表明，VDE实现了显著加速，同时视觉质量损失极小：使Flux加速3.22倍，在Qwen-Image上达到LPIPS 0.069，优于最佳基线52.2%。
原文摘要 · Abstract (English)
Though rectified flow models have achieved remarkable performance in image, video, and 3D generation, their practical deployments are challenged by slow inference speeds. Prior acceleration methods reuse cached features from previous steps, which neglects the growing mismatch between static caches and the evolving input, leading to reduced output fidelity. This work proposes Velocity Decomposition and Estimation (VDE), a training-free acceleration method that shifts the paradigm from caching-and-reusing to decomposing-and-estimating. Specifically, VDE decomposes the model's velocity into components parallel and orthogonal to the input, exploiting their temporal predictability and directional stability for precise, input-adaptive estimation. To prevent error accumulation, it periodically anchors the model's state via full forward passes. Extensive experiments on image and video generation tasks demonstrate that VDE achieves substantial acceleration with minimal loss in visual quality. Notably, VDE accelerates Flux by 3.22 times and achieves an LPIPS of 0.069 on Qwen-Image, outperforming the best baseline with a 52.2% reduction.
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。