CASIA-SURF
收藏资源简介:
CASIA-SURF是一个大规模多模态面部反欺骗基准数据集,由中国科学院自动化研究所创建。该数据集包含1000个不同性别、年龄和室内环境的主题,每个主题有1个真实视频和6个欺骗视频,总计21000个视频。数据集提供RGB、深度和红外三种模态的数据,旨在通过多模态信息提高面部识别系统的安全性。创建过程中,使用Intel RealSense SR300相机在多种室内环境下采集数据,确保了数据的真实性和多样性。CASIA-SURF数据集主要用于面部反欺骗技术的研究,特别是在面部支付、解锁等安全敏感场景中的应用。
CASIA-SURF is a large-scale multimodal facial anti-spoofing benchmark dataset developed by the Institute of Automation, Chinese Academy of Sciences. It encompasses 1000 subjects with varying genders, ages and indoor settings, where each subject is associated with 1 genuine video and 6 spoofing videos, resulting in a total of 21,000 videos. The dataset provides data in three modalities: RGB, depth and infrared, with the goal of enhancing the security of facial recognition systems through multimodal information. During its construction, data was collected using Intel RealSense SR300 cameras across diverse indoor environments, ensuring the authenticity and diversity of the dataset. The CASIA-SURF dataset is primarily utilized for research on facial anti-spoofing technologies, particularly for applications in security-sensitive scenarios such as facial payment and device unlocking.

- 1CASIA-SURF: A Large-scale Multi-modal Benchmark for Face Anti-spoofing中国科学院自动化研究所 · 2020年



