Raghavendra Ramachandra

Presentation Attack Detection Methods for Face Recognition Systems: A Comprehensive Survey

The vulnerability of face recognition systems to presentation attacks (also known as direct attacks or spoof attacks) has received a great deal of interest from the biometric community. The rapid evolution of face recognition systems into real-time applications has raised new concerns about their ability to resist presentation attacks, particularly in unattended application scenarios such as automated border control. The goal of a presentation attack is to subvert the face recognition system by presenting a facial biometric artifact. Popular face biometric artifacts include a printed photo, the electronic display of a facial photo, replaying video using an electronic display, and 3D face masks. These have demonstrated a high security risk for state-of-the-art face recognition systems. However, several presentation attack detection (PAD) algorithms (also known as countermeasures or antispoofing methods) have been proposed that can automatically detect and mitigate such targeted attacks. The goal of this survey is to present a systematic overview of the existing work on face presentation attack detection that has been carried out. This paper describes the various aspects of face presentation attacks, including different types of face artifacts, state-of-the-art PAD algorithms and an overview of the respective research labs working in this domain, vulnerability assessments and performance evaluation metrics, the outcomes of competitions, the availability of public databases for benchmarking new PAD algorithms in a reproducible manner, and finally a summary of the relevant international standardization in this field. Furthermore, we discuss the open challenges and future work that need to be addressed in this evolving field of biometrics.

Biometric Systems under Morphing Attacks: Assessment of Morphing Techniques and Vulnerability Reporting

With the widespread deployment of biometric recognition systems, the interest in attacking these systems is increasing. One of the easiest ways to circumvent a biometric recognition system are so-called presentation attacks, in which artefacts are presented to the sensor to either impersonate another subject or avoid being recognised. In the recent past, the vulnerabilities of biometric systems to so-called morphing attacks have been unveiled. In such attacks, biometric samples of multiple subjects are merged in the signal or feature domain, in order to allow a successful verification of all contributing subjects against the morphed identity. Being a recent area of research, there is to date no standardised manner to evaluate the vulnerability of biometric systems to these attacks. Hence, it is not yet possible to establish a common benchmark between different morph detection algorithms. In this paper, we tackle this issue proposing new metrics for vulnerability reporting, which build upon our joint experience in researching this challenging attack scenario. In addition, recommendations on the assessment of morphing techniques and morphing detection metrics are given.

Assessing face image quality for smartphone based face recognition system

In recent years, the popularity of smartphones has increased massively as a personal and authentication device. Face based biometrics is being used to secure the device and control access to several different services via smartphones such as payment gateways etc. Thus, to maintain the reliability and to obtain better verification performance, there is a need to adopt the standards recommended for face sample quality. In this paper, we present an evaluation of face image quality assessment using well-established ISO standards on the images collected using smartphones. In this work, we constructed a new database of 101 individuals with 22 frontal face images with different facial pose angles, illumination and at five different distances between the subject and the mobile device. We evaluate the existing quality metrics and further propose a new quality metric based on vertical edge density that can robustly estimate the pose variations and improves the quality estimation of a face image. The proposed method is evaluated for reliable estimation of the quality for smartphone face biometrics.

Biometric Authentication Protocols on Smartphones: An Overview

As biometric authentication methods become more and more common in daily life we have seen an increased interest and developments of more convenient and secure authentication methods for online services. With biometrics enabled smartphones the cost associated with deploying biometric systems is removed and the opportunity for use in new applications is opened. However, with biometrics there are several challenges in terms of security and privacy that must be addressed. In this work we look at two emerging authentication protocols, FIDO Universal Authentication Framework and Biometric Open Protocol Standard, and analyze their security and privacy features from a biometrics perspective.

Multi-biometric template protection based on bloom filters

Abstract Biometric verification systems are currently being deployed in numerous large-scale and everyday applications. It is hence of the utmost importance to protect the privacy of the enrolled subjects. Biometric template protection schemes are designed to protect biometric reference data in an irreversible and unlinkable manner, while maintaining key system properties like the accuracy or the speed. In past years, template protection schemes based on Bloom filters have been introduced and applied to various biometric characteristics. While the irreversibility and unlinkability of Bloom filter-based protection schemes have been shown, their application to any given unprotected template is not straightforward. In this article we present a methodology for the estimation of the main parameters of such schemes, based on a statistical analysis of the unprotected templates. Furthermore, in order to increase verification accuracy and privacy protection, a general approach for a protected weighted feature level fusion is proposed. In order to avoid biased results, the soundness of the estimation methodologies is confirmed for face, iris, fingerprint and fingervein over two totally different sets of publicly available databases. In addition, we show how the weighted feature level fusion preserves the accuracy of the unprotected score level fusion, while it adds privacy protection to the system.

Multi-biometric template protection — A security analysis of binarized statistical features for bloom filters on smartphones

In recent years, we have seen huge growth of biometric systems incorporated in devices such as smartphones and security is one of the major concerns. In this work a multi-biometric template protected system is proposed, based on Bloom filers and binarized statistical image features (BSIF). Features are extracted from face and both periocular regions and templates protected using Bloom filters. Score level fusion is applied to increase recognition accuracy. The system is tested on a database, consisting of 94 subjects, of images collected with smart phones. A comparison between unprotected and protected templates in the system shows the feasibility of the template protection method with observed Genuine-Match-Rate (GMR) of 95.95% for unprotected templates and 91.61% at a False-Match-Rate (FMR) of 0.01%. Irreversibility and unlinkability of the system is analysed based on a recently published security evaluation framework.

Collaborative Representation of Statistically Independent Filters’ Response: An Application to Face Recognition Under Illicit Drug Abuse Alterations

Face biometrics is widely deployed in many security and surveillance applications that demand a secure and reliable authentication service. The performance of face recognition systems is primarily based on the analysis of texture and geometric variation of the face. Continuous and extensive consumption of illicit drugs will significantly result in deformation of both texture and geometric characteristics of a face and thus, impose additional challenges on accurately identifying the subjects who abuse drugs. This work proposes a novel scheme to improve robustness of face recognition system to address the variations caused by the prolonged use of illicit drugs. The proposed scheme is based on the collaborative representation of statistically independent filters whose responses are computed on the face images captured before and after substance (or drug) abuse. Extensive experiments are carried out on the publicly available Illicit Drug Abuse Database (DAD) comprised of face images from 100 subjects. The obtained results indicate better performance of the proposed scheme when compared with six different state-of-the-art approaches including a commercial face recognition system.

Feature level fused templates for multi-biometric system on smartphones

This work examines feature level fusion for protected biometric templates in a multi-biometric authentication system for smartphones. The modalities incorporated by the system are face and the left-right periocular region. The fusion methods considered are concatenation of the templates obtained from the three modalities, and combining the three templates using a simple XOR operation by varying the amount of overlap between them up to 100%. The impact on performance from applying the feature level fusion methods are evaluated on a moderate sized dataset consisting of images from 73 subjects, captured using a Samsung Galaxy S5. We show that the biometric performance can be improved in most of the cases by employing the fusion methods when compared to the performance of each individual modality while not compromising the security level provided by template protection schemes.

Multi-biometric Template Protection on Smartphones: An Approach Based on Binarized Statistical Features and Bloom Filters.

Widespread use of biometric systems on smartphones raises the need to evaluate the feasibility of protecting biometric templates stored on such devices to preserve privacy. To this extent, we propose a method for securing multiple biometric templates on smartphones, applying the concepts of Bloom filters along with binarized statistical image features descriptor. The proposed multi-biometric template system is first evaluated on a dataset of 94 subjects captured with Samsung S5 and then tested in a real-life access control scenario. The recognition performance of the protected system based on the facial characteristic and the two periocular regions is observed equally good as the baseline performance of unprotected biometric system. The observed Genuine-Match-Rate (GMR) of \(91.61\%\) at a False-Match-Rate (FMR) of \(0.01\%\) indicates the robustness and applicability of the proposed system in everyday authentication scenario. The reliability of the system is further tested by engaging disjoint subset of users, who were tasked to use the proposed system in their daily activities for a number of days. Obtained results indicate the robustness of the proposed system to preserve user privacy while not compromising the inherent authentication accuracy without protected templates.