GenVidBench
收藏资源简介:
GenVidBench是由华为诺亚方舟实验室创建的一个大规模、高质量的AI生成视频检测数据集。该数据集包含143,131条视频,涵盖8种最先进的AI视频生成器生成的视频,并提供了丰富的语义标签,如对象类别、动作和位置。数据集通过跨源和跨生成器的设计,确保了训练集和测试集之间的多样性,避免了内容过于相似的问题。GenVidBench旨在解决AI生成视频检测中的挑战,帮助研究人员开发和评估更有效的检测模型,应用于防止虚假信息的传播和网络安全威胁的应对。
GenVidBench is a large-scale, high-quality AI-generated video detection dataset developed by Huawei Noah's Ark Lab. This dataset contains 143,131 video clips, covering content generated by 8 state-of-the-art AI video generators, and is annotated with rich semantic tags including object categories, actions, and spatial locations. Adopting a cross-source and cross-generator design, the dataset ensures diversity between the training and test sets, thus avoiding the problem of overly similar content. GenVidBench is intended to address the challenges in AI-generated video detection, enabling researchers to develop and evaluate more effective detection models, which can be applied to curbing the spread of disinformation and addressing cybersecurity threats.

- 1GenVidBench: A Challenging Benchmark for Detecting AI-Generated Video华为诺亚方舟实验室 · 2025年



