FloodDiffusion：面向流式动作生成的定制化扩散强制
FloodDiffusion: Tailored Diffusion Forcing for Streaming Motion Generation
论文揭示的扩散强制改造要点（双向注意力、下三角时间调度器、连续时变条件注入）可直接迁移到视频等流式时序生成任务，为实时、变条件驱动的序列生成提供可复用的训练范式。
原文依据 we adopt a diffusion forcing framework to model this time-series generation task under time-varying control events
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文提出 FloodDiffusion，一个用于文本驱动的流式人体动作生成的新框架。给定随时间变化的文本提示，FloodDiffusion 能够以实时延迟生成与文本对齐、无缝衔接的动作序列。与依赖逐块生成或带扩散头的自回归模型的现有方法不同，我们采用扩散强制框架，在时变控制事件下对该时间序列生成任务进行建模。我们发现，直接套用原始扩散强制（如为视频模型所提出的版本）无法对真实动作分布进行建模。我们证明，为保证对输出分布的建模，原始扩散强制必须做如下定制：(i) 使用双向注意力而非因果注意力进行训练；(ii) 采用下三角时间调度器而非随机调度器；(iii) 采用连续时变的方式引入文本条件。借助这些改进，我们首次证明基于扩散强制的框架在流式动作生成任务上达到最优性能，在 HumanML3D 基准上取得 0.057 的弗雷歇初始距离。模型、代码与权重均已公开。
英文原文摘要 1187 字符
We present FloodDiffusion, a new framework for text-driven, streaming human motion generation. Given time-varying text prompts, FloodDiffusion generates text-aligned, seamless motion sequences with real-time latency.Unlike existing methods that rely on chunk-by-chunk or auto-regressive model with diffusion head, we adopt a diffusion forcing framework to model this time-series generation task under time-varying control events.We find that a straightforward implementation of vanilla diffusion forcing (as proposed for video models) fails to model real motion distributions. We demonstrate that to guarantee modeling the output distribution, the vanilla diffusion forcing must be tailored to: (i) train with a bi-directional attention instead of casual attention; (ii) implement a lower triangular time scheduler instead of a random one; (iii) utilize a continues time-varying way to introduce text conditioning.With these improvements, we demonstrate in the first time that the diffusion forcing-based framework achieves state-of-the-art performance on the streaming motion generation task, reaching an FID of 0.057 on the HumanML3D benchmark. Models, code, and weights are available.