论文标题
低成本增强的安全面部识别的立体声摄像头识别
Low cost enhanced security face recognition with stereo cameras
论文作者
论文摘要
本文探讨了面部识别替代方案，该替代方法旨在为大多数识别体系结构中的当前安全漏洞做出贡献。由于缺乏深度信息，当前市场中当前的低成本面部验证软件可以被面部印刷图片所欺骗。呈现的软件在立体声设置的帮助下创建了面部深度图，比传统识别程序提供了更高的安全性。对人的身份和面部深度图的分析是通过深层卷积神经网络处理的，提供了安全的低成本实时面部身份验证方法。
This article explores a face recognition alternative which seeks to contribute to resolve current security vulnerabilities in most recognition architectures. Current low cost facial authentication software in the market can be fooled by a printed picture of a face due to the lack of depth information. The presented software creates a depth map of the face with the help of a stereo setup, offering a higher level of security than traditional recognition programs. Analysis of the person's identity and facial depth map are processed through deep convolutional neural networks, providing a secure low cost real-time face authentication method.
