Valid

GB/T 41815.1-2022Information technology - Biometric presentation attack detection - Part 1: Framework (English PDF)

Preview PDF

This is a limited preview

Buy now to download the full PDF (24 pages)

Need quotation for your management?

Get a formal quotation for GB/T 41815.1-2022 in 30 minutes.

Issued by

State Administration for Market Regulation, China National Standardization Administration

Level / Type

National · Recommended

Issue date

October 12, 2022

Implementation date

May 1, 2023

Scope

GB/T 41815.1-2022 (Information technology - Biometric presentation attack detection - Part 1: Framework) is available as an English-translated PDF.

GB/T 41815.1-2022 is the Chinese standard "Information technology - Biometric presentation attack detection - Part 1: Framework". Its scope clause reads: This document establishes terms and definitions commonly used in the specification, characterization and evaluation of presentation attack detection methods.

The following are outside the scope of this document: - standardization of specific presentation attack detection methods; - detailed information about countermeasures (such as anti-spoofing techniques), algorithms or sensors; - system-level security or vulnerability assessment of biometric systems.

The attacks considered in this document are those that take place at the sensor during the presentation and collection of biometric characteristics. Other attacks are outside the scope of this document.

Found what you need?

Document preview — GB/T 41815.1-2022

National Standard of the People's Republic of China

ICS: 35.240.01
Classification: L 67

Issued by: State Administration for Market Regulation, China National Standardization Administration

Foreword
Introduction
1 Scope
2 Normative references
3 Terms and definitions
4 Abbreviations
5 Characterization of presentation attacks
5.1 General
5.2 Presentation attack instruments
6 Framework of presentation attack detection methods
6.1 Types of presentation attack detection
6.2 Role of challenge/response
6.3 Presentation attack detection process
6.4 Presentation attack detection in biometric system architecture
7 Means of deterring impostor presentation attacks in biometric systems
Bibliography

Foreword

This document was drafted in accordance with the rules given in GB/T 1.1-2020 Directives for standardization - Part 1: Rules for the structure and drafting of standardizing documents.

This document is Part 1 of GB/T 41815 Information technology - Biometric presentation attack detection. The following parts of GB/T 41815 have been published:

- Part 1: Framework;

- Part 2: Data formats.

This document is identical to ISO/IEC 30107-1:2016 Information technology - Biometric presentation attack detection - Part 1: Framework.

Attention is drawn to the possibility that some of the elements of this document may be the subject of patent rights. The issuing body of this document shall not be held responsible for identifying patent rights.

This document was proposed by and is under the jurisdiction of the National Information Technology Standardization Technical Committee (SAC/TC 28).

Drafting organizations of this document: China Electronics Standardization Institute, Lenovo Zhongtian Technology Co., Ltd., Ping An Technology (Shanghai) Co., Ltd., Shanghai SenseTime Intelligent Technology Co., Ltd., Xiamen Yicheng Information Technology Co., Ltd., Shenzhen CESI Information Technology Co., Ltd., Beijing Megvii Technology Co., Ltd., Beijing Wanlihong Technology Co., Ltd., Beijing Eyecool Intelligent Technology Co., Ltd., Beijing IrisKing Technology Co., Ltd., Institute of Automation, Chinese Academy of Sciences, Chipone Technology (Beijing) Co., Ltd., CloudWalk Technology Group Co., Ltd., Tianjin Zhongke Hongxing Technology Co., Ltd., Guangzhou Mailun Information Technology Co., Ltd., Hangzhou Jinglianwen Technology Co., Ltd., China Academy of Information and Communications Technology, Newland Digital Technology Co., Ltd., Ant Group Co., Ltd., Huizhou University, Tianfu (Dongguan) Standard Technology Co., Ltd., Xi'an Jiaotong University, Beijing University of Posts and Telecommunications, GRG Banking Equipment Co., Ltd., Shanghai Yitu Network Technology Co., Ltd., Beijing Pensees Technology Co., Ltd., Shenzhen Tencent Computer Systems Co., Ltd., State Grid Blockchain Technology (Beijing) Co., Ltd.

Introduction

Biometric technologies recognize individuals by their physiological and behavioural characteristics and are therefore often used as components of security systems. Biometric technologies assist security systems in attempting to recognize friend or foe, and can also attempt to recognize persons unknown to the system.

Since biometric technologies came into use, the possibility of attacks against them has been recognized, and it is therefore necessary to take measures to detect and defend against subversive recognition attempts and presentation attacks. Attacks on the intended function of biometric technologies can occur at any time and at any point within a security system, and can be launched by any actor, whether a person inside or outside the system. The scope of this document is limited to techniques that automatically detect presentation attacks carried out by the biometric capture subject at the time of presentation and capture of the relevant biometric characteristics, i.e. "presentation attack detection" (PAD) methods.

The individual who is the biometric capture subject presents a potential security risk at the time of data capture, which limits the use of biometrics in applications that are not supervised by the system owner or its agent, for example remote capture over untrusted networks. For example, guidelines on electronic authentication do not recommend the use of biometric technologies as an authentication factor for this purpose. In unattended applications, such as remote identity verification over open networks, automated presentation attack detection methods can be applied to mitigate the risk of attack. Standards, best practices and independently evaluated technologies can improve the security of all systems that use biometric technologies, whether data capture is supervised or unsupervised, including systems that use biometric technologies to secure online transactions.

As with biometric technologies, PAD technologies can also make errors, including false positives and false negatives: a false positive wrongly classifies a normal presentation as an attack, thereby impairing the efficiency of the system; a false negative wrongly classifies a presentation attack as a normal presentation and fails to prevent the security breach. The choice of a specific PAD implementation will therefore depend on the specific application and on the trade-off between security and efficiency.

GB/T 41815 Information technology - Biometric presentation attack detection specifies the framework, data formats and testing-related content for presentation attack detection in the field of biometrics, to facilitate the design of presentation attack detection functions and the evaluation of their capability. GB/T 41815 is intended to consist of 3 parts.

- Part 1: Framework. Its purpose is to establish the overall framework related to presentation attack detection in biometric systems; it is applicable to the design and use of biometric systems.

- Part 2: Data formats. Its purpose is to specify the data formats related to presentation attack detection; it is applicable to data exchange between biometric systems.

- Part 3: Testing and reporting. Its purpose is to clarify the factors to be considered in evaluating presentation attack detection capability, together with the evaluation principles and methods; it is applicable to the analysis and evaluation of the presentation attack detection capability of biometric systems.

Standardizing the subjects related to presentation attack detection first requires the presentation attack detection framework to be defined as a whole, to guide the classification of attacks and the definition of instruments; secondly, in order to exchange and share the relevant data elements, clear and uniform data formats need to be defined; finally, the evaluation of presentation attack detection capability requires the test objects, test environment and test metrics to be specified, to guide a scientific and objective evaluation of presentation attack detection capability.

1 Scope

This document establishes terms and definitions commonly used in the specification, characterization and evaluation of presentation attack detection methods.

The following are outside the scope of this document:

- standardization of specific presentation attack detection methods;

- detailed information about countermeasures (such as anti-spoofing techniques), algorithms or sensors;

- system-level security or vulnerability assessment of biometric systems.

The attacks considered in this document are those that take place at the sensor during the presentation and collection of biometric characteristics.

Other attacks are outside the scope of this document.

2 Normative references

The following documents, through normative references in the text, constitute indispensable provisions of this document. For dated references, only the edition corresponding to that date applies to this document. For undated references, the latest edition (including all amendments) applies to this document.

GB/T 5271.37-2021 Information technology - Vocabulary - Part 37: Biometrics (ISO/IEC 2382-37:2017 1), MOD)

Note: There is no technical difference between the content of GB/T 5271.37-2021 that is referenced and the content of ISO/IEC 2382-37:2012 that is referenced.

1) The latest edition of ISO/IEC 2382-37 is the edition revised in 2022; the content relevant to this document has not been technically revised.

3 Terms and definitions

For the purposes of this document, the terms and definitions given in GB/T 5271.37-2021 and the following apply.

3.1 artefact

Artificial object or representation presenting a copy of biometric characteristics or a synthetic biometric modality.

3.2 liveness

State or quality of vital signs evidently produced by anatomical characteristics, involuntary reactions (or physiological functions) or voluntary reactions (or subject behaviours).

Example 1: Absorption of light by skin and blood is an anatomical characteristic.

Example 2: The reaction of the iris to light and heart activity (pulse) are involuntary reactions (also known as physiological functions).

Example 3: Making a gesture by squeezing the fingers together and a biometric presentation made in response to an instruction prompt are voluntary reactions (also known as subject behaviours).

3.3 liveness detection

Measurement and analysis of anatomical characteristics or of involuntary or voluntary reactions, in order to determine whether a captured biometric sample comes from a living person present at the point of capture.

Note: Liveness detection methods are a subset of presentation attack detection methods.

3.4 normal presentation

Interaction of the biometric capture subject with the biometric data capture subsystem in the manner intended by the policy of the biometric system.

Note: When referring to "normal presentation", the term "normal" is equivalent to "routine". Any presentation that is not an attack is regarded as a "normal presentation".

3.5 presentation attack

Presentation to the biometric data capture subsystem with the goal of interfering with the normal operation of the biometric system.

Note 1: Presentation attacks can be implemented in various ways, such as artefact attacks, mutilation attacks, replay attacks, etc.

Note 2: Presentation attacks may have various goals, such as impersonation or avoiding recognition.

Note 3: Biometric systems may have difficulty distinguishing between presentation attacks aimed at interfering with system operation and presentation attacks that do not conform to the system requirements.

3.6 presentation attack detection

Automated determination of a presentation attack.

Note: Presentation attack detection cannot infer the intent of the subject. In fact, it is difficult to derive differences in the intent of the subject from the data capture process or from the captured samples.

3.7 presentation attack instrument

Biometric characteristic or object used in a presentation attack.

Note: The set of presentation attack instruments includes not only artefacts but also lifeless biometric characteristics used for attack (such as those taken from cadavers) or altered biometric characteristics (such as altered fingerprints).

4 Abbreviations

The following abbreviations apply to this document.

PAD Presentation Attack Detection

PAI Presentation Attack Instrument

5 Characterization of presentation attacks

5.1 General

Although any attacker may attack a biometric system at any point of processing, this document focuses on biometric-based attacks on the data capture subsystem carried out by biometric capture subjects attempting to subvert the normal operation of the biometric system. Attacks carried out by other attackers and at other points of the system have already been considered in other documents (Bibliography [2]). This document does not deal with preventing the data capture subsystem (including the sensor itself) from being modified, replaced or removed, nor with protecting the communication between the data capture subsystem and other subsystems.

Several general attacks against biometric systems are shown in Figure 1. This document focuses only on the attacks indicated by arrow "1", in which a biometric characteristic or PAI is presented to a normally operating data capture sensor of the biometric system.

Figure 1 - Examples of attack points in a biometric system (stages: data capture, within the scope of GB/T 41815; signal processing; comparison; decision; data storage. Attack points: 1 presentation attack at the sensor; 2 biometric sample tampering; 3 signal processing unit override; 4 probe tampering; 5 comparator override; 6 database override and tampering; 7 biometric reference tampering; 8 score tampering; 9 decision override)

Presentation attacks may be carried out by two types of subversive biometric capture subject. One is the biometric impostor, i.e. a subversive biometric capture subject who intends to be recognized as another individual (not himself/herself); the other is the biometric concealer, i.e. a subversive biometric capture subject who intends to avoid being recognized as an enrolled user of the system.

A biometric impostor may attack in two different ways. In one, the subversive biometric data subject intends the biometric system to recognize him/her as a specific individual enrolled in the system; in the other, the subversive biometric data subject intends the biometric system to recognize him/her as any enrolled but unspecified individual.

In contrast, a biometric concealer attempts to conceal his/her own biometric characteristics by imitating the characteristics of an unenrolled individual (for example by using an artefact, or by disguising or modifying the original biometric characteristics).

5.2 Presentation attack instruments

Objects or characteristics used in presentation attacks are called PAIs. Attacks using PAIs at the sensor generally fall into one of two categories: artificial characteristics or human-based characteristics. Note that there is a third category of other natural cases, such as PAIs based on animals and plants.

In addition, the two terms normal and abnormal are used in this subclause, in particular in Table 1, because the meaning of these two terms relates only to the interaction of the subject with the sensor, and such interaction is difficult to measure and code objectively, so it does not affect PAD coding. An example of abnormal interaction is touching the fingerprint capture device with the side of the finger instead of the whole fingerprint.

It should be noted that a detected attack may be caused by accessibility or usability problems of the subject rather than by an attempt to attack the system.

The further subdivided PAI categories are shown in the third row (bottom layer) of Figure 2. Table 1 gives examples of each specific PAI type in the bottom layer of Figure 2. The figure can be used to describe a specific PAI by using the adjective in the second column (of Table 1) followed by the word in the first column. For example, a part of a cadaver is a "lifeless, human PAI".

Figure 2 - Types of presentation attack (presentation attack instrument: artificial - complete, partial; human - lifeless, altered, abnormal, coerced, normal; other natural)

Table 1 - Examples of artificial and human presentation attack instruments

Artificial, complete: silicone finger film, face video.

Artificial, partial: glue on the finger, sunglasses, artificial/patterned contact lenses, non-permanent make-up.

Human, lifeless: part of a cadaver, severed finger/hand.

Human, altered: mutilation, surgical exchange of finger and toe prints.

Human, abnormal: abnormal facial expression, fingertip or side of the finger.

Human, coerced (a): unconscious, under duress.

Human, normal: no impersonation attempt.

a Not all "coerced" presentations can be detected. Some modalities may be able to measure traces of coercion, such as voice emotion analysis, extreme pulse rate, facial emotion analysis (e.g. fear).

Remaining clauses in the full document

6 Framework of presentation attack detection methods
7 Means of deterring impostor presentation attacks in biometric systems
Bibliography

......
This preview omits tables, figures, formulas and parts of the technical clauses. The complete document — 24 pages — is available in the English PDF.

Referenced standards

Normative references

GB/T

Cited by

GB/T 41815.2-2022Information technology - Biometric presentation attack detection - Part 2: Data formats

How to Download GB/T 41815.1-2022

1
Add to cart
Click "Download PDF" on this page and choose "Buy this standard". You can add more standards before checkout.
2
Checkout
Enter your email and billing details. Payment is processed securely by Stripe (cards or bank transfer). Prefer bank details? Request a quotation (min. order $300).
3
Instant delivery (0–9 sec)
Delivery is automatic: within seconds of payment you'll receive an email with a secure download link. The link stays valid for 72 hours.
4
Invoice included
A tax invoice is attached to the confirmation email. Need a custom invoice? Contact us.