Dynamic Multi-Process Liveness Detection for Identity Fraud Prevention

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing living body judgement technologies are insufficient in preventing identity fraud, as they can be easily bypassed by using pre-prepared images that are designed to be recognized as living bodies by these technologies.

Innovation Solution

A judgement method that selects multiple living body judgement processes, determines their order of execution, and requires the subject to perform specific actions for each process, making it difficult for malicious individuals to prepare images that would pass all selected processes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If multiple living body judgement processes are selected and executed in a determined order, then the reliability of identity verification is improved, but the device complexity increases

Engineering Contradiction:
Improveidentity verification reliabilityVSAvoidjudgement system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The living body judgement system is divided into multiple independent judgement processes (first, second, third, and fourth processes) that can be selectively executed. Each process focuses on specific facial features (e.g., eye blinking, mouth movement, face orientation) and can be independently configured and evaluated, allowing the system to achieve high reliability through cumulative assessment while maintaining modular complexity management.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system dynamically selects and orders multiple judgement processes based on predefined conditions and requirements. The execution sequence and selection of specific processes can be adjusted according to the verification needs, allowing the system to adapt its complexity level while maintaining reliable verification. This dynamic configuration enables the system to scale from simple to complex verification scenarios as needed.

Inventive Principle:
Principle #15Dynamics

2Measurement precision

If specific actions are required for each living body judgement process, then the ability to detect fraudulent images is improved, but the ease of operation deteriorates

Engineering Contradiction:
Improvefraud detection precisionVSAvoidoperation convenience
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The system provides instructive information to the subject in advance, clearly indicating what actions are required for each judgement process (e.g., blink eyes, move mouth, turn face). This preliminary guidance ensures that subjects understand the expected behaviors, enabling precise detection of living body characteristics while simplifying the interaction process. The advance instruction reduces confusion and improves both detection accuracy and user experience.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system provides real-time feedback during the judgement processes, informing subjects of their current status and required actions. This feedback mechanism ensures that subjects can perform the correct actions to demonstrate they are living beings, while the system accurately captures and evaluates these actions. The interactive feedback loop enhances detection precision by ensuring proper action execution while maintaining ease of operation through clear communication.

Inventive Principle:
Principle #23Feedback

3Object-affected harmful factors

If multiple judgement processes are executed sequentially, then the difficulty of preparing fraudulent images increases, but the loss of time increases

Engineering Contradiction:
Improveidentity fraud difficultyVSAvoidverification time
Core Design Contradiction:
Object-affected harmful factorsVSLoss of time

Solution Approach 1:

The system executes multiple judgement processes in a periodic or sequential manner, where each process builds upon the previous one. This structured approach ensures that fraudulent images must satisfy multiple independent criteria (different facial actions, orientations, and temporal patterns) rather than a single check. The periodic execution of diverse judgement processes significantly increases the difficulty of creating convincing fake images while managing verification time through efficient process ordering and parallel evaluation where possible.

Inventive Principle:
Principle #19Periodic action

Data Source

PatentUS12223774B2Judgement method, judgement device, and judgement program
Publication Date: 2025.02.11 SWALLOW INCUBATE CO LTD
  • US12223774B2 patent drawing
  • US12223774B2 patent drawing
  • US12223774B2 patent drawing

AI summary

A judgement method includes: selecting two or more living body judgement processes among a plurality of living body judgement processes; determining an order of executing the two or more selected living body judgement processes; executing the two or more selected living body judgement processes in the determined order; executing, in each of the two or more selected living body judgement processes: a processing of showing a subject an action required; a processing of acquiring a facial image containing a face of the subject when acting in accordance with the shown action; and a processing of judging whether the subject is of living body based on a feature of a section of the face contained in the facial image; and judging that the subject is of living body when judgement results obtained from the two or more selected living body judgement processes satisfy a predetermined condition.