一个算法对齐所有：基于 Rectified Flow 的结构对齐联合推理
One Algorithm to Align Them All
提出基于 Rectified Flow 的联合传输方法，可在任意结构化潜变量空间上实现成对样本的结构对齐生成，替代 SDS 用于 3D 生成时速度提升数量级，为图像/视频/3D 的配对生成提供可复用框架。
原文依据 Our approach can be built on top of an arbitrary Rectified Flow model operating on the structured latent space.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论我们提出一种新的多模态算法，用于在 Rectified Flow 模型下对成对的结构对齐样本进行联合推理。尽管现有的一些方法提出了相互依赖的生成过程，但它们并未从结构对齐的角度看待联合生成问题。近期工作使用分数蒸馏采样（SDS）来生成对齐的 3D 模型，但 SDS 已知耗时、易出现模式崩溃，且通常产生卡通化结果。相比之下，我们提出的方法依赖于样本空间中一个片段的联合传输，从而在推理时实现更快的计算。我们的方法可以构建在任意作用于结构化潜变量空间的 Rectified Flow 模型之上。我们使用最优基线展示了该方法在图像、视频和 3D 形状生成领域的适用性，并分别与基于编辑和基于联合推理的竞争方法进行了评估。我们证明了用该方法获得的样本对具有高度的结构对齐性，且样本具有高视觉质量。我们的方法提升了图像和视频生成流程的最优性能。对于 3D 生成，它能够以数量级更快的速度展示出可比的质量。
英文原文摘要 1265 字符
We suggest a new multi-modal algorithm for joint inference of paired structurally aligned samples with Rectified Flow models. While some existing methods propose a codependent generation process, they do not view the problem of joint generation from a structural alignment perspective. Recent work uses Score Distillation Sampling to generate aligned 3D models, but SDS is known to be time-consuming, prone to mode collapse, and often provides cartoonish results. By contrast, our suggested approach relies on the joint transport of a segment in the sample space, yielding faster computation at inference time. Our approach can be built on top of an arbitrary Rectified Flow model operating on the structured latent space. We show the applicability of our method to the domains of image, video, and 3D shape generation using state-of-the-art baselines and evaluate it against both editing-based and joint inference-based competing approaches. We demonstrate a high degree of structural alignment for the sample pairs obtained with our method and a high visual quality of the samples. Our method improves the state-of-the-art for image and video generation pipelines. For 3D generation, it is able to show comparable quality while working orders of magnitude faster.