arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2510.08073cs.CVcs.LG
收藏

物理驱动的时空建模用于AI生成视频检测

Physics-Driven Spatiotemporal Modeling for AI-Generated Video Detection

South China University of Technology(华南理工大学);University of Science and Technology of China(中国科学技术大学);Key Laboratory of Big Data and Intelligent Robot, Ministry of Education(教育部大数据与智能机器人重点实验室);Pazhou Lab(琶洲实验室);University of Melbourne(墨尔本大学);Hunan University(湖南大学);Hong Kong Baptist University(香港 Baptist大学)

Shuhai Zhang, ZiHao Lian, Jiahao Yang, Daiyuan Li, Guoxuan Pang, Feng Liu, Bo Han, Shutao Li, Mingkui Tan

AI总结:

提出基于概率流守恒的物理驱动AI生成视频检测范式,通过归一化时空梯度(NSG)统计量捕捉物理异常,结合预训练扩散模型估计NSG,并利用最大均值差异(MMD)进行检测,在Recall和F1-Score上分别提升16.00%和10.75%。

摘要:

AI生成的视频已实现近乎完美的视觉真实感(如Sora),迫切需要可靠的检测机制。然而,检测此类视频在建模高维时空动态和识别违反物理规律的细微异常方面面临重大挑战。本文提出首个基于概率流守恒原理的物理驱动AI生成视频检测范式。具体而言,我们提出一种称为归一化时空梯度(NSG)的统计量,该统计量量化空间概率梯度与时间密度变化之比,明确捕捉与自然视频动态的偏差。利用预训练的扩散模型,我们通过空间梯度近似和运动感知时间建模开发了NSG估计器,无需复杂的运动分解,同时保持物理约束。在此基础上,我们提出基于NSG的视频检测方法(NSG-VD),该方法计算测试视频与真实视频NSG特征之间的最大均值差异(MMD)作为检测指标。最后,我们推导了真实视频与生成视频之间NSG特征距离的上界,证明由于分布偏移,生成视频表现出放大的差异。大量实验证实,NSG-VD在Recall和F1-Score上分别比最先进的基线方法高出16.00%和10.75%,验证了NSG-VD的优越性能。源代码可在该 https URL 获取。

英文摘要:

AI-generated videos have achieved near-perfect visual realism (e.g., Sora), urgently necessitating reliable detection mechanisms. However, detecting such videos faces significant challenges in modeling high-dimensional spatiotemporal dynamics and identifying subtle anomalies that violate physical laws. In this paper, we propose the first physics-driven AI-generated video detection paradigm based on probability flow conservation principles. Specifically, we propose a statistic called Normalized Spatiotemporal Gradient (NSG), which quantifies the ratio of spatial probability gradients to temporal density changes, explicitly capturing deviations from natural video dynamics. Leveraging pre-trained diffusion models, we develop an NSG estimator through spatial gradients approximation and motion-aware temporal modeling without complex motion decomposition while preserving physical constraints. Building on this, we propose an NSG-based video detection method (NSG-VD) that computes the Maximum Mean Discrepancy (MMD) between NSG features of the test and real videos as a detection metric. Last, we derive an upper bound of NSG feature distances between real and generated videos, proving that generated videos exhibit amplified discrepancies due to distributional shifts. Extensive experiments confirm that NSG-VD outperforms state-of-the-art baselines by 16.00% in Recall and 10.75% in F1-Score, validating the superior performance of NSG-VD. The source code is available at https://github.com/ZSHsh98/NSG-VD.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑