Anti-spoofing face recognition using a metasurface-based snapshot hyperspectral image sensor

Abstract

Modern face recognition systems usually combine RGB, depth, and infrared cameras to do face antispoofing, but they are still not robust enough to unknown 3D high-quality mask attack. In our work, we developed a snapshot hyperspectral image sensor based on metasurface nanostructures to obtain the high-precision hyperspectral information of the detected face, and we built a practical antispoofing face recognition system using our new sensor. Experiments show that our sensor can reconstruct the reflectance spectrum of human skin, and this spectral information captured by our sensor can be quite effective and robust to identify spoof faces. We attack our system with several types of spoof faces, and our system reaches 97.98% accuracy in real-world testing scenes.

A novel cost-effective vivo facial detection technique based on snapshot spectral imaging sensor

“…In the range of visible light, there is a characteristic spectrum of human skin at 500-650nm [9,12] , which can be used as one of the anti-spoofing criteria. Therefore, the transmission spectrum center wavelength of CIS is designed between 400nm and 750nm, and it has high reconstruction performance in this range.…”

“…Traditional face anti-spoofing (FAS) system mostly based on RGB-classification algorithm or near infrared (NIR) images [4,7,8] , which is used to identify the face in printing photos and screens, but vulnerable to 3D disguise masks [9] . To defend against these attacks, the 3D face recognition techniques that acquire 3D images are developed such as RGBD camera [10] and structured light camera [11] .…”

Random broadband filters based on combination of metasurface and multilayer thin films for hyperspectral imaging

“…Simultaneously, it is noteworthy that the metasurface based filters could provide spectral channels much more than the number of units. This strategy not only effectively minimizes the size of the spectrometer but also paves the way for developing more compact spectroscopy equipment [41][42][43]. Generally, metasurface filters with intricate unit elements are required to generate a multitude of broadband random spectral responses, which are instrumental in achieving high spectral resolution.…”

Optical Angle Barcodes Enabled by a Pixelated Metasurface for Absolute Micro‐Angle Measurement

“…[32] By assigning different block pattern design to individual pixel, pixelated metasurface has found wide applications in spectral visualization, such as molecular fingerprint identification, spectral imaging, facial recognition, or nano-painting. [33][34][35][36] Inspired by the ability of pixelated metasurface to convert optical information into visual images, we propose the concept of optical angle barcodes. The pixelated structure is employed to encode the angular information contained in the customized scattering field into an image.…”