Abstract:Today, deep learning represents the most popular and successful form of machine learning. Deep learning has revolutionised the field of pattern recognition, including biometric recognition. Biometric systems utilising deep learning have been shown to achieve auspicious recognition accuracy, surpassing human performance. Apart from said breakthrough advances in terms of biometric performance, the use of deep learning was reported to impact different covariates of biometrics such as algorithmic fairness, vulnerability to attacks, or template protection. Technologies of biometric template protection are designed to enable a secure and privacy-preserving deployment of biometrics. In the recent past, deep learning techniques have been frequently applied in biometric template protection systems for various purposes. This work provides an overview of how advances in deep learning take influence on the field of biometric template protection. The interrelation between improved biometric performance rates and security in biometric template protection is elaborated. Further, the use of deep learning for obtaining feature representations that are suitable for biometric template protection is discussed. Novel methods that apply deep learning to achieve various goals of biometric template protection are surveyed along with deep learning-based attacks.

What problem does this paper attempt to address?

The main issue this paper attempts to address is exploring how deep learning technology impacts the field of biometric template protection and providing a comprehensive overview of research in this area. Specifically, the paper focuses on the following aspects: 1. **Background of Biometric Template Protection**: - Biometric recognition systems have made significant progress in terms of accuracy, but they have also raised concerns regarding privacy and security. - Traditional encryption methods are not suitable for biometric data because biometric data exhibits intra-class variability. 2. **Application of Deep Learning in Biometric Template Protection**: - Deep learning technology can bring significant benefits in various processing steps such as feature extraction, feature type conversion, and feature alignment. - The paper discusses how to use deep learning to generate feature representations suitable for biometric template protection and how these methods can improve the privacy and security of the system. 3. **Relationship Between Performance and Privacy Protection**: - Deep learning has improved the accuracy of biometric recognition, allowing the system to operate at a lower False Match Rate (FMR), thereby enhancing system security. - However, the enhanced generalization ability of deep learning systems may also make them more susceptible to certain attacks. 4. **Feature Type Conversion**: - The paper details how to convert different types of feature representations (such as real-valued vectors, integer vectors, and binary vectors) to adapt to different biometric template protection schemes. - These conversion steps may lead to information loss, but by appropriately selecting parameters, biometric performance can be maintained. 5. **Application of Deep Networks in Template Protection**: - Recent research attempts to embed biometric template protection directly into the deep learning feature extraction process by introducing randomness to enhance protection. - The paper discusses the specific implementation methods of these approaches and their potential vulnerabilities. In summary, this paper aims to provide researchers in the field of biometric template protection with a comprehensive perspective, introducing the latest advancements and applications of deep learning technology in this area.

Deep Learning in the Field of Biometric Template Protection: An Overview

Deep Learning for Biometrics

Exploring Human Biometrics: A Focus on Security Concerns and Deep Neural Networks

Biometric Template Protection for Neural-Network-based Face Recognition Systems: A Survey of Methods and Evaluation Techniques

Biometrics recognition using deep learning: a survey

ECG Biometric Recognition: Template-Free Approaches Based on Deep Learning

Biometrics for Forensic Identification in Web Applications and Social Platforms Using Deep Learning

Unlocking Security: The Significant Role of Machine Learning and Deep Learning Techniques in Fingerprint Recognition Systems

Research on Bi-mode Biometrics Based on Deep Learning

AttackNet: Enhancing Biometric Security via Tailored Convolutional Neural Network Architectures for Liveness Detection

Deep Secure Encoding: An Application to Face Recognition

Presentation Attack detection using Wavelet Transform and Deep Residual Neural Net

BioDeepHash: Mapping Biometrics into a Stable Code

AdvBiom: Adversarial Attacks on Biometric Matchers

Beyond Identity: What Information Is Stored in Biometric Face Templates?

Biometric lock with facial recognition implemented with deep learning techniques

A Survey on Homomorphic Encryption for Biometrics Template Security Based on Machine Learning Models

Deep Learning based Fingerprint Presentation Attack Detection: A Comprehensive Survey

Review on Deep Learning-Based Face Analysis

Privacy-Preserving Eye-tracking Using Deep Learning

Advancements in Face Recognition: From Feature Extraction to Deep Learning Models and Integrated Biometric Solutions