发表机构
Shanghai Jiao Tong University; Cardiff University（上海交通大学; 卡迪夫大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究物理上合理的视频生成问题，提出PhysAgent反射式智能体框架，通过闭环设计及物理控制API，改善参数控制，实现复杂轨迹等，实验证明其能生成更合理视频，有更好提示对齐及泛化效果。
AI 中文摘要
基于物理的视频生成的最新进展利用物理模拟作为物理先验来指导视频合成以获得物理上合理的结果。模拟过程由物理规范控制，通常由视觉语言模型一次性生成。这种一次性预测往往无法将用户意图准确转化为可执行的模拟，特别是对于细粒度物体动力学、复杂运动轨迹和时间结构化交互。本文提出了PhysAgent，一个反射式智能体框架，它在物理程序生成、物理模拟、特定阶段验证和目标程序修复之间形成闭环。除了改善对耦合物理参数的控制，我们的框架使智能体能够通过将每个物理程序视为可执行假设来逐步实现复杂轨迹、多阶段交互和精确事件结果。此外，我们设计了一组物理控制API以支持更稳定和复杂的运动行为。大量实验表明，PhysAgent能生成更符合物理原理的视频，实现更好的提示对齐，并在不同物理场景中更有效地泛化。
英文摘要
Recent advances in physics-grounded video generation leverage physics simulation as a physical prior to guide video synthesis toward physically plausible outcomes. The simulation process is controlled by physical specifications, which are typically generated by a vision-language model in a single pass. Such one-shot prediction often fails to accurately translate user intent into executable simulations, particularly for fine-grained object dynamics, complex motion trajectories, and temporally structured interactions. In this paper, we propose PhysAgent, a reflective agentic framework that closes the loop among physical program generation, physics simulation, stage-specific verification, and targeted program repair. Beyond improving the control of coupled physical parameters, our framework enables the agent to progressively realize complex trajectories, multi-stage interactions, and precise event outcomes by treating each physical program as an executable hypothesis. In addition, we design a set of physics-control APIs to support more stable and complex motion behaviors. Extensive experiments demonstrate that PhysAgent produces more physically plausible videos, achieves better prompt alignment, and generalizes more effectively across diverse physical scenarios.
CommentsFor project page, see https://iapple233.github.io/PhysAgent
