ActivityForensics：面向视频中被操纵活动定位的综合基准
ActivityForensics: A Comprehensive Benchmark for Localizing Manipulated Activity in Videos
为视频取证提供了「活动级」伪造的定位基准与跨域/开放世界评测协议，并给出基于扩散特征正则化的基线 TADiff；做视频生成检测、时序定位或伪造溯源的研究者可直接在其协议上评测，或借鉴其用扩散模型暴露生成伪影痕迹的思路。
原文依据 we introduce ActivityForensics, the first large-scale benchmark for localizing manipulated activity in videos
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论时序伪造定位旨在从时间维度上识别视频中被篡改的片段。现有基准大多聚焦于外观层面的伪造，例如换脸与物体移除。然而，视频生成技术的最新进展催生了活动层面的伪造：这类伪造通过修改人物动作来扭曲事件语义，产生极具欺骗性的伪造内容，严重损害媒体真实性与公众信任。为解决这一问题，我们提出 ActivityForensics，这是首个用于定位视频中被操纵活动的大规模基准。它包含超过 6K 个伪造视频片段，这些片段被无缝融入视频上下文，呈现出高度视觉一致性，使人眼几乎无法将其与真实内容区分。我们进一步提出时序伪影扩散器（Temporal Artifact Diffuser，TADiff），这是一个简单而有效的基线，通过基于扩散的特征正则化器暴露伪影线索。基于 ActivityForensics，我们提出涵盖域内、跨域与开放世界设定的综合评测协议，并对一系列最优伪造定位器进行基准评测，以推动未来研究。数据集与代码见 https://activityforensics.github.io。
英文原文摘要 1266 字符
Temporal forgery localization aims to temporally identify manipulated segments in videos. Most existing benchmarks focus on appearance-level forgeries, such as face swapping and object removal. However, recent advances in video generation have driven the emergence of activity-level forgeries that modify human actions to distort event semantics, resulting in highly deceptive forgeries that critically undermine media authenticity and public trust. To overcome this issue, we introduce ActivityForensics, the first large-scale benchmark for localizing manipulated activity in videos. It contains over 6K forged video segments that are seamlessly blended into the video context, rendering high visual consistency that makes them almost indistinguishable from authentic content to the human eye. We further propose Temporal Artifact Diffuser (TADiff), a simple yet effective baseline that exposes artifact cues through a diffusion-based feature regularizer. Based on ActivityForensics, we introduce comprehensive evaluation protocols covering intra-domain, cross-domain, and open-world settings, and benchmark a wide range of state-of-the-art forgery localizers to facilitate future research. The dataset and code are available at https://activityforensics.github.io.