Vol.:(0123456789) 1 3 Journal of Banking and Financial Technology https://doi.org/10.1007/s42786-020-00016-z ORIGINAL ARTICLE Comprehensive survey on cancelable biometrics with novel case study on fnger dorsal template protection Avantika Singh 1  · Ashish Arora 2  · Gaurav Jaswal 3  · Aditya Nigam 1 Received: 4 July 2019 / Accepted: 4 April 2020 © Institute for Development and Research in Banking Technology 2020 Abstract In order to use biometric based authentication in banking domain, it is vital to secure biometric databases from unauthor- ized access. As we all know a biometric trait is immutable, thus if it is stolen it is lost forever. Therefore, securing biometric databases is a paramount concern in today’s digital World. In this paper, we have presented a comprehensive study of bio- metric template protection schemes majorly focusing on cancelable biometric techniques. Moreover, we have proposed a cancelable fnger dorsal template generation network trained on trait specifc features without using any pre-trained network framework. The highly discriminative features learned from our network are further hashed using BioHashing technique to generate cancelable fnger dorsal templates. The experimental results are evaluated on two benchmark publicly available fnger knuckle databases: PolyU FKP and PolyU Contactless FKI. We have also performed in depth security analysis of the proposed framework in terms of invertibility, revocability and unlinkability. Keywords Biometric template protection · Cancelable biometrics · Cancelable fnger dorsal templates · BioHashing 1 Introduction In today’s 21st century, modern banks have been enthusias- tically adopting biometric based authentication for various purposes. Now banks are using biometrics for developing new generation identifcation control mechanism that can combat fraud, make transactions hassle-free and can provide better customer satisfaction. It has been estimated that by the end of year 2020, about 1.9 billion bank customers around the World will use biometrics for various applications like withdrawing cash from the ATMs, proving their identity to customer care centers via telephone authentication (Citi Bank uses voice biometrics ), authenticating a mobile bank app and many more. In all these applications to gain public confdence and trust, it is important to secure them. Due to the advancement of technology in today’s World, biometric traits can be captured covertly and can be stolen, thus possess a serious security threat for biometric based authentication systems. Conventional data encryption algo- rithms like SHA, MD5 are not applicable on biometric data due to the presence of large variance in it. As we all know, conventional biometric systems always perform fuzzy com- parisons by setting a decision threshold between imposter and genuine score distributions. In order to handle this issue of stolen biometrics and large variance in biometric signals, researchers have proposed two biometric template protection schemes namely (i) biometric cryptosystems, (ii) cancelable biometrics. In both of these techniques, original biometric template is never stored in the database, thus prevents from database attacks. Here, in this paper we will discuss in detail about various template protection schemes majorly focusing on cancelable biometrics along with its recent performance evaluation metrics. Moreover, we will also present a case study on generating cancelable fnger dorsal templates using state-of-the art BioHashing technique. 1.1 Biometric cryptosystems In case of biometric cryptosystems(BCSs), biometric dependent information which is commonly known as helper data is stored in the database. Here, the biometric verifcation is performed by validating the keys generated or * Avantika Singh d16027@students.iitmandi.ac.in 1 Indian Institute of Technology Mandi, Mandi, HP, India 2 Indian Institute of Technology Dharwad, Dharwad, Karnataka, India 3 Indian Institute of Technology Delhi, Delhi, India