Biometric Registration Neural Network Encoding Update

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional biometric information registration systems using artificial neural networks face challenges in maintaining and updating the networks without initial biometric input, making it difficult to apply rapidly developing deep learning technologies effectively.

Innovation Solution

A biometric information registration method and apparatus that constructs an artificial neural network by connecting an encoding network and a feature point extraction network, allowing for continuous updates using converted information instead of re-receiving biometric data, thereby renewing the biometric information template.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If conventional biometric information registration systems are used, then the system can perform initial biometric registration, but the artificial neural network cannot be updated without re-receiving initial biometric input

Engineering Contradiction:

ImproveAbility to update neural networkVSAvoidRequirement for re-inputting biometric data

Core Design Contradiction:

Adaptability or versatilityVSEase of operation

Solution Approach 1:

The system performs preliminary encoding of biometric information into converted information during the initial registration phase. This converted information is stored and can be used for future updates without requiring re-input of the original biometric data, thereby enabling easy network updates while maintaining operational convenience

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces converted information as an intermediary between the original biometric information and the artificial neural network. This intermediary representation allows the network to be updated and maintained without directly handling the original biometric data, solving the contradiction between update capability and operational ease

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If the artificial neural network is updated frequently to incorporate rapidly developing deep learning technology, then the system can maintain high security and accuracy, but the complexity of maintaining and supplementing the network increases

Engineering Contradiction:

ImproveSecurity and accuracy of biometric systemVSAvoidComplexity of maintaining and supplementing neural network

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system segments the biometric information processing into distinct components: the original biometric information, the converted information (encoding result), and the artificial neural network. This segmentation allows the network to be updated independently using the stored converted information, reducing the complexity of maintenance while maintaining high security and accuracy through frequent updates

Inventive Principle:

Principle #1Segmentation

3Measurement precision

If the artificial neural network is designed to accept biometric information as initial input, then the system can perform accurate feature extraction, but it becomes difficult to maintain and supplement the network without the original biometric input

Engineering Contradiction:

ImproveAccuracy of feature extractionVSAvoidAbility to update without original input

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The system performs preliminary encoding of biometric information into a compact converted representation before feeding it to the neural network. This preliminary action preserves the essential features needed for accurate extraction while creating a self-contained update mechanism that doesn't require the original biometric input, thereby resolving the contradiction between extraction accuracy and update adaptability

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS12189737B2Biometric information registration method and biometric information registration device for performing same

Publication Date: 2025.01.07 ALCHERA INC

AI summary

The embodiments disclosed herein relate to a biometric information registration apparatus and method using an artificial neural network. According to an embodiment, there are disclosed a biometric information registration apparatus and method that register biometric information by deriving converted information and a biometric information template based on an artificial neural network formed by successively connecting an encoding network for extracting security information and a feature point extraction network for extracting feature points to each other and renew the biometric information template using the converted information when updating the artificial neural network.