Multispectral Biometrics System Framework: Application to Presentation Attack Detection
This work addresses security vulnerabilities in biometric systems for applications like access control, though it is incremental as it builds on existing multispectral methods with a broader spectral range.
The authors tackled the problem of presentation attack detection in biometrics by developing a multispectral framework that captures data across visible to long-wave-infrared wavelengths, and they analyzed the data using a deep-learning classifier to evaluate each spectral band's effectiveness in distinguishing live from fake samples.
In this work, we present a general framework for building a biometrics system capable of capturing multispectral data from a series of sensors synchronized with active illumination sources. The framework unifies the system design for different biometric modalities and its realization on face, finger and iris data is described in detail. To the best of our knowledge, the presented design is the first to employ such a diverse set of electromagnetic spectrum bands, ranging from visible to long-wave-infrared wavelengths, and is capable of acquiring large volumes of data in seconds. Having performed a series of data collections, we run a comprehensive analysis on the captured data using a deep-learning classifier for presentation attack detection. Our study follows a data-centric approach attempting to highlight the strengths and weaknesses of each spectral band at distinguishing live from fake samples.