Biometric Attack Testing With Robotic Finger Placement

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current biometric recognition testing methods are inefficient and inaccurate due to the stochastic nature of machine learning algorithms, requiring extensive human resources and resulting in long test periods and high error variances, making it difficult to reproduce results.

Innovation Solution

A biometric recognition attack test method using a mechanical arm and biometric feature objects with conductive parts to simulate user interactions, allowing for automated testing of biometric recognition devices by placing and pressing these objects into recognition areas, thereby reducing manual errors and improving test efficiency and accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual testing with test volunteers is used, then biometric recognition capability can be tested, but human resources are consumed, test period is long, and error variances are large

Engineering Contradiction:
Improvetest accuracyVSAvoidtest efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent uses artificial fingerprints as copies of real fingerprints to perform biometric recognition testing. These artificial fingerprints replicate the essential characteristics of real fingerprints without requiring human volunteers, thereby eliminating human resource consumption and reducing error variances while maintaining test validity

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent replaces the manual mechanical process of human volunteers providing fingerprints with an automated system that uses artificial fingerprints and a drive unit to present fingerprint images to the recognition device, eliminating manual operation errors and improving test efficiency

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If multiple attempts are made due to stochastic machine learning algorithms, then test reliability improves, but human resources increase and test period extends

Engineering Contradiction:
Improvetest reliabilityVSAvoidtest period
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

Artificial fingerprints can be precisely replicated and reused multiple times without degradation, allowing repeated testing attempts with identical parameters. This eliminates the need to recruit new human volunteers for each attempt, reducing time loss while maintaining test reliability through consistent, reproducible test samples

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The artificial fingerprints are prepared in advance with precise control over their characteristics, allowing multiple test attempts to be conducted with predetermined conditions. This preliminary preparation eliminates the variability inherent in manual collection from human volunteers, ensuring consistent reliability across multiple attempts without extending the test period

Inventive Principle:
Principle #10Preliminary action

3Ease of operation

If manual testing operations are performed, then flexibility in testing is maintained, but error variances between operations increase

Engineering Contradiction:
Improvetesting flexibilityVSAvoidtest consistency
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent replaces manual testing operations with an automated drive unit that precisely controls the presentation of artificial fingerprints to the biometric recognition device. This mechanical substitution eliminates human operational variability while maintaining flexibility through programmable control, ensuring consistent test results without sacrificing ease of operation

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The artificial fingerprints are designed to be self-presenting through the drive unit mechanism, which automatically positions and presents them to the recognition device without requiring manual handling. This self-service approach eliminates inter-operator variability while maintaining operational simplicity through automated control

Inventive Principle:
Principle #25Self-service

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

The method significantly reduces human resource consumption, minimizes test duration, and enhances the accuracy and reproducibility of biometric recognition testing by using a mechanical arm to simulate user interactions with biometric recognition devices.

Implementation Method 1

the biometric feature object includes a feature attachment part to which biometric features are attached and a conductive part

Methodology Applied
Scientific EffectElectrical conduction: Conduction (electrical)

Data Source

PatentEP3940562B1Biometric recognition attack test methods, apparatuses, and devices
Publication Date: 2023.12.20 ALIPAY (HANGZHOU) INFORMATION TECH CO LTD
  • EP3940562B1 patent drawingFigure 1~2A
  • EP3940562B1 patent drawingFigure 2B
  • EP3940562B1 patent drawingFigure 2C

AI summary

Embodiments of the present specification disclose a biometric recognition attack test method, apparatus, and device. The method includes: obtaining a biometric feature object needed for performing a biometric recognition attack on a biometric recognition device, where the biometric feature object includes a feature attachment part to which biometric features are attached and a conductive part; controlling a mechanical arm to place the biometric feature object in a recognition area of the biometric recognition device, and controlling the mechanical arm to press the biometric feature object to trigger the biometric feature object to input the biometric feature in the feature attachment part into the biometric recognition device through the conductive part, to perform a biometric recognition attack test; and obtaining an attack test result corresponding to the biometric feature object, and determining, based on the attack test result corresponding to the biometric feature object, a test result of the biometric recognition attack performed on the biometric recognition device.