MoReGen：面向基于代码的文本到视频合成的多智能体运动推理引擎
MoReGen: Multi-Agent Motion-Reasoning Engine for Code-based Text-to-Video Synthesis
把物理仿真器与渲染器接入多智能体LLM流水线，在代码域生成可复现视频，并用物体轨迹对应作为直接评测指标；对做视频生成的人，这提供了一条绕开像素级扩散、直接保证牛顿运动一致性的可复用路线与配套基准。
原文依据 we propose object-trajectory correspondence as a direct evaluation metric
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论尽管文本到视频（T2V）生成在照片级真实感方面已取得显著进展，但生成忠实遵循物理原理、与意图对齐的视频仍是一项核心挑战。本文系统研究牛顿运动可控的文本到视频生成与评测，强调物理精度与运动连贯性。我们提出 MoReGen，一个运动感知、物理落地的 T2V 框架，它集成多智能体大语言模型（LLM）、物理仿真器与渲染器，在代码域中从文本提示生成可复现、物理准确的视频。为定量评估物理有效性，我们提出物体-轨迹对应关系作为直接评测指标，并给出 MoReSet——一个包含 1,275 段人工标注视频的基准，覆盖九类牛顿现象，并附带场景描述、时空关系与真值轨迹。利用 MoReSet，我们对现有 T2V 模型开展实验，通过我们的 MoRe 指标与现有基于物理的评测器评估其物理有效性。结果表明，最优模型难以维持物理有效性，而 MoReGen 为物理连贯的视频合成确立了一个有原则的方向。
英文原文摘要 1254 字符
While text-to-video (T2V) generation has achieved remarkable progress in photorealism, generating intent-aligned videos that faithfully obey physics principles remains a core challenge. In this work, we systematically study Newtonian motion-controlled text-to-video generation and evaluation, emphasizing physical precision and motion coherence. We introduce MoReGen, a motion-aware, physics-grounded T2V framework that integrates multi-agent large language models (LLMs), physics simulators, and renderers to generate reproducible, physically accurate videos from text prompts in the code domain. To quantitatively assess physical validity, we propose object-trajectory correspondence as a direct evaluation metric and present MoReSet, a benchmark of 1,275 human-annotated videos spanning nine classes of Newtonian phenomena with scene descriptions, spatiotemporal relations, and ground-truth trajectories. Using MoReSet, we conduct experiments on existing T2V models, evaluating their physical validity through both our MoRe metrics and existing physics-based evaluators. Our results reveal that state-of-the-art models struggle to maintain physical validity, while MoReGen establishes a principled direction toward physically coherent video synthesis.