AxonData/3D_paper_mask_attack_dataset_for_Liveness

3D Paper Mask Attack数据集专注于3D体积纸攻击，包含鼻子、肩膀和额头等元素，适用于PAD级别1和2的活体检测测试。数据集包括使用各种移动设备捕获的视频，并采用了主动活体检测技术。数据集具有40多名参与者，涵盖了高加索人、黑人和亚洲人种，视频在iOS和Android手机上捕获，每段攻击视频大约7秒。数据集还包括缩放阶段以模拟主动活体检测，并提供了多样化的场景，如添加围巾、眼镜和连帽衫等体积元素，以及使用低端和高端设备捕获的特定攻击场景和动作。数据集仍在数据收集阶段，欢迎反馈和建议。数据集适用于活体检测、iBeta活体测试、反欺骗、生物识别认证、机器学习和深度学习等领域。

The 3D Paper Mask Attack Dataset focuses on 3D volume-based paper attacks, incorporating elements such as the nose, shoulders, and forehead. These attacks are designed to be advanced and are useful for both PAD level 1 and level 2 liveness tests. The dataset includes videos captured using various mobile devices and incorporates active liveness detection techniques. Key features include over 40 participants with a balanced representation of ethnicities, video capture on iOS and Android phones, active liveness detection with zoom-in and zoom-out phases, diverse scenarios including volume-based elements like scarves, glasses, and hoodies, and specific paper types used for attacks. The dataset is still in the data collection phase and welcomes feedback and requests to incorporate additional features or specific requirements. Potential use cases include liveness detection, iBeta liveness testing, anti-spoofing, biometric authentication, and training and evaluating machine learning and deep learning models.