基于流式力控制的流式视频生成

摘要

我们提出了StreamForce，一个流式视频生成框架，能够通过连续力输入实现物理基础控制。不同于以往需要为不同力类型训练独立模型、假设固定力或依赖非因果处理的视频模型，StreamForce是一个因果统一模型，能够即时且连贯地响应局部和全局的时变力。为此，我们设计了一种统一的力表示作为控制信号，并开发了一套力可控视频生成的蒸馏流程。该模型结合了自回归效率与力响应能力，维持稳定的光度与动态真实感。StreamForce在单GPU上可达每秒16.6帧，在力控制一致性与运动真实感方面均达到最优性能。项目网站：https://neu-vi.github.io/StreamForce/

English

We introduce StreamForce, a streaming video generation framework that enables physically grounded control through continuous force inputs. Unlike prior video models that train separate models for different force types, assume fixed forces, or rely on non-causal processing, StreamForce is a causal and unified model that responds instantly and coherently to both local and global, time-varying forces. To achieve this, we design a unified force representation as a control signal and develop a distillation pipeline for force-controllable video generation. Our model combines autoregressive efficiency with force responsiveness, sustaining stable photometric and dynamic realism. StreamForce runs at up to 16.6 FPS on a single GPU, achieving state-of-the-art performance in both force adherence and motion realism. Project website: https://neu-vi.github.io/StreamForce/