基于流匹配的统一无人数约束文本到动作生成
Unified Number-Free Text-to-Motion Generation Via Flow Matching
把 flow matching 用在多人动作生成上，用统一潜空间桥接异构动作数据集、用金字塔分辨率条件化降低开销、用半噪声联合概率路径缓解自回归误差累积，这套「单次先验生成 + 多次反应生成」的分解思路可迁移到多人交互视频生成与人体动作预测。
原文依据 UMF decomposes the number-free motion generation into a single-pass motion prior generation stage and multi-pass reaction generation stages.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论生成模型在固定智能体数量的动作合成上表现优异，但难以泛化到智能体数量可变的情形。现有方法基于有限的领域特定数据，采用自回归模型递归生成动作，存在效率低和误差累积的问题。我们提出统一动作流（Unified Motion Flow, UMF），由金字塔动作流（Pyramid Motion Flow, P-Flow）与半噪声动作流（Semi-Noise Motion Flow, S-Flow）组成。UMF 将无人数约束的动作生成分解为单次动作先验生成阶段与多次反应生成阶段。具体而言，UMF 利用统一潜空间弥合异构动作数据集之间的分布差异，从而实现有效的统一训练。在动作先验生成方面，P-Flow 在不同噪声水平的条件下作用于层次化分辨率，从而降低计算开销。在反应生成方面，S-Flow 学习一条联合概率路径，自适应地执行反应变换与上下文重建，从而缓解误差累积。大量实验结果与用户研究表明，UMF 作为通用模型在文本驱动的多人动作生成上是有效的。项目主页：https://githubhgh.github.io/umf/。
英文原文摘要 1228 字符
Generative models excel at motion synthesis for a fixed number of agents but struggle to generalize with variable agents. Based on limited, domain-specific data, existing methods employ autoregressive models to generate motion recursively, which suffer from inefficiency and error accumulation. We propose Unified Motion Flow (UMF), which consists of Pyramid Motion Flow (P-Flow) and Semi-Noise Motion Flow (S-Flow). UMF decomposes the number-free motion generation into a single-pass motion prior generation stage and multi-pass reaction generation stages. Specifically, UMF utilizes a unified latent space to bridge the distribution gap between heterogeneous motion datasets, enabling effective unified training. For motion prior generation, P-Flow operates on hierarchical resolutions conditioned on different noise levels, thereby mitigating computational overheads. For reaction generation, S-Flow learns a joint probabilistic path that adaptively performs reaction transformation and context reconstruction, alleviating error accumulation. Extensive results and user studies demonstrate UMF's effectiveness as a generalist model for multi-person motion generation from text. Project page: https://githubhgh.github.io/umf/.