[论文] Streaming Video Generation with Streaming Force Control

论文概要

研究领域: CV
作者: Hanhui Wang, Yiming Xie, Haiwen Feng
发布时间: 2025-06-11
arXiv: 2506.08644

中文摘要

本文提出StreamForce，一种流式视频生成框架，通过连续力输入实现物理层面的可控生成。与此前针对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方法不同，StreamForce是一个因果统一模型，能够对局部与全局的时变力输入做出即时且连贯的响应。为此，我们设计了统一力表征作为控制信号，并开发了力可控视频生成的蒸馏管线。该模型结合自回归效率与力响应能力，保持稳定的真实感与动态真实感。StreamForce在单GPU上可达16.6 FPS，在力遵循度与运动真实感方面均达到最优性能。

原文摘要

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 m...

自动采集于 2026-06-09

#论文 #arXiv #CV #小凯