Client-Side Biometric Authentication for Privacy Protection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for user identity verification in online services require explicit user consent for biometric data collection and processing, leading to complex compliance issues and high computational costs due to server-side analysis, while also being vulnerable to fraud if consent is denied.

Innovation Solution

A method that analyzes biometric data on the client device, generating user-specific machine learning templates locally without sharing sensitive data, allowing continuous authentication by evaluating biometric patterns like typing and mouse movements, and sending only a score to the server for verification, thus reducing compliance burdens and computational load on the server.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If biometric data are collected and sent to the back-end for server-side analysis, then user identity verification can be performed, but privacy protection becomes complex and compliance management becomes burdensome

Engineering Contradiction:
Improveuser identity verificationVSAvoidprivacy protection and compliance management
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent extracts the biometric data processing function from the server and relocates it to the client device. The client device collects biometric data, generates machine learning templates, and performs authentication analysis locally, while only sending non-sensitive authentication results to the server. This extraction eliminates the need for complex privacy protection and compliance management on the server side while maintaining reliable user identity verification.

Inventive Principle:
Principle #2Taking out (Extraction)

2Extent of automation

If server-side analysis is used for biometric data processing, then centralized control is maintained, but computational cost increases significantly

Engineering Contradiction:
Improvecentralized controlVSAvoidcomputational cost
Core Design Contradiction:
Extent of automationVSPower

Solution Approach 1:

The patent segments the authentication system into client-side processing and server-side verification. The computationally intensive tasks of biometric data collection, machine learning template generation, and authentication analysis are performed on the client device. The server only performs lightweight verification of authentication results. This segmentation dramatically reduces server computational cost while maintaining automated authentication control.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The client device performs self-service authentication by locally generating machine learning templates from biometric data and conducting authentication analysis without requiring server resources. The client device autonomously manages its own authentication process, sending only minimal authentication results to the server for record-keeping and session management.

Inventive Principle:
Principle #25Self-service

3Object-affected harmful factors

If explicit user consent is required for biometric data collection, then privacy protection is ensured, but user experience becomes complex and compliance management becomes burdensome

Engineering Contradiction:
Improveprivacy protectionVSAvoiduser experience and compliance management
Core Design Contradiction:
Object-affected harmful factorsVSEase of operation

Solution Approach 1:

The patent extracts the privacy-sensitive biometric data processing from the server to the client device. Since processing occurs locally on the client, no explicit user consent is required for server processing, and privacy protection is maintained through local data handling. The client device generates and stores machine learning templates locally, eliminating the need for complex consent management while ensuring privacy.

Inventive Principle:
Principle #2Taking out (Extraction)

4Measurement precision

If machine learning templates are created for each individual user, then accurate authentication is achieved, but computational resources required increase with the number of users

Engineering Contradiction:
Improveauthentication accuracyVSAvoidcomputational resources
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

Each client device performs self-service authentication by generating and storing its own machine learning templates locally. The client device autonomously creates templates from user biometric data and maintains them in local storage, eliminating the need for server-side template management. This approach achieves accurate authentication for each user while distributing the computational resource burden across individual devices rather than concentrating it on the server.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP4123483A1Method for confirming the identity of a user in a browsing session of an online service
Publication Date: 2023.01.25 CLEAFY SPA
  • EP4123483A1 patent drawingFigure 1
  • EP4123483A1 patent drawingFigure 2
  • EP4123483A1 patent drawingFigure 3

AI summary

Method for confirming the identity of a user (U) in a browsing session of an online service, comprising the steps of: a) providing a web server (1) in which an online service resides, in communication with a client device (2) provided with a user interface (21); b) providing a database (11) associated with the web server (1) in which a plurality of data relating to one or more users (U, U', U") registered to the online service are stored; c) providing a script (22) residing in the client device (2); d) identifying via script (22) each browsing session on the online service and associating it with a user (U) registered to the online service when the latter performs authentication; e) collecting via script (22) biometric data generated by said at least one user interface (21) and associating them with the user (U) when authenticated; f) generating via script (22) machine learning templates as a result of processing the biometric data; g) storing the biometric data and the machine learning templates locally in the client device (2); h) generating a score associated with the user (U) as a result of processing via script (22) new biometric data collected on said at least one user interface (21) as a function of the machine learning templates generated in step f); i) sending the score to the web server (1); 1) verifying the identity of the authenticated user (U) as a result of processing the score by means of a security algorithm (12) residing in the web server (1).