Distance entropy as an information measure for binary biometric representation

Authors:
Yi Cheng Feng;Pong Chi Yuen;Meng-Hui Lim
Affiliations:
Department of Computer Science, Hong Kong Baptist University, Hong Kong;Department of Computer Science, Hong Kong Baptist University, Hong Kong;Department of Computer Science, Hong Kong Baptist University, Hong Kong
Venue:
CCBR'12 Proceedings of the 7th Chinese conference on Biometric Recognition
Year:
2012

Citing 10
Cited 0

Elements of information theory

Elements of information theory
Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection

IEEE Transactions on Pattern Analysis and Machine Intelligence
Random Multispace Quantization as an Analytic Mechanism for BioHashing of Biometric and Random Identity Inputs

IEEE Transactions on Pattern Analysis and Machine Intelligence
Cancellable biometrics and annotations on BioHash

Pattern Recognition
Orientation selection using modified FCM for competitive code-based palmprint recognition

Pattern Recognition
Towards a measure of biometric feature information

Pattern Analysis & Applications
A hybrid biometric cryptosystem for securing fingerprint minutiae templates

Pattern Recognition Letters
A hybrid approach for generating secure and discriminating face template

IEEE Transactions on Information Forensics and Security
How to Measure Biometric Information?

ICPR '10 Proceedings of the 2010 20th International Conference on Pattern Recognition
An efficient dynamic reliability-dependent bit allocation for biometric discretization

Pattern Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

To uphold the security of individuals, analyzing the amount of information in binary biometric representation is highly essential. While Shannon entropy is a measure to quantify the expected value of information in the binary representation, it does not account the extent to which every binary representation could distinctively identify a person in a population. Hence, it does not appropriately quantify the hardness of obtaining a close approximation of the user's biometric template if one maliciously leverages the population distribution. To resolve this, relative entropy has been used to measure information of user distribution with reference to the population distribution. However, existing relative-entropy estimation techniques that are based on statistical methods in the Euclidean space cannot be directly extended to the Hamming space. Therefore, we put forward a new entropy measure known as distance entropy and its estimation technique to quantify the information in binary biometric representation more effectively with respect to the discrimination power of the binary representation.