Facial Micro-Movement Authentication for Attack-Resistant User Authentication

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing biometric authentication mechanisms, such as facial recognition, are vulnerable to attacks and lack sufficient resistance, as they rely on immutable biometric features that can be easily mimicked.

Innovation Solution

Authentication is enhanced by recognizing facial micro-movements during a prompted facial gesture, utilizing user-specific reference samples to verify the authenticity of the user.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If facial recognition is used for authentication, then the authentication process is seamless and easy to use, but the security resistance against attacks is insufficient

Engineering Contradiction:
Improveauthentication easeVSAvoidsecurity resistance
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The patent transitions from static facial recognition to dynamic facial micro-movement analysis. The system captures and analyzes real-time facial muscle movements (micro-expressions) that occur naturally during authentication, making the biometric data dynamic and difficult to replicate with static masks or casts.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the authentication parameter from static facial geometry to dynamic facial muscle movement patterns. By measuring subtle changes in facial feature positions over time during a prompted gesture, the system obtains a new biometric parameter that is much harder to spoof while maintaining ease of use.

Inventive Principle:
Principle #35Parameter changes

2Ease of operation

If biometric features are used for authentication, then the authentication is seamless, but the features are immutable and cannot be changed if compromised

Engineering Contradiction:
Improveauthentication seamlessVSAvoidbiometric changeability
Core Design Contradiction:
Ease of operationVSAdaptability or versatility

Solution Approach 1:

The system uses dynamic facial movements rather than static biometric features. Since the authentication is based on movement patterns during a prompted gesture rather than immutable physical traits, the system gains adaptability - if compromised, the gesture prompt can be changed while maintaining the same authentication mechanism.

Inventive Principle:
Principle #15Dynamics

3Reliability

If facial micro-movements are used for authentication, then the resistance against attacks is improved, but the measurement and detection difficulty increases

Engineering Contradiction:
Improveattack resistanceVSAvoidmicro-movement detection
Core Design Contradiction:
ReliabilityVSDifficulty of detecting and measuring

Solution Approach 1:

The system prompts the user to perform a specific facial gesture before capturing the micro-movements. This preliminary action standardizes the authentication process and ensures that the facial muscles are actively engaged, making the micro-movements more detectable and consistent across different users.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent adds the time dimension to facial recognition by analyzing facial micro-movements over time. Instead of a single static image, the system captures a sequence of facial states during a gesture, transforming the problem from 2D spatial recognition to 3D spatio-temporal analysis, which improves detectability of subtle movements.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentEP3991087B1Method, apparatus and computer program for authenticating a user
Publication Date: 2025.10.29 SONY GROUP CORP
  • EP3991087B1 patent drawingFigure 1a~1b
  • EP3991087B1 patent drawingFigure 1c~2

AI summary

Examples relate to a method, an apparatus and a computer program for authenticating a user. The method comprises prompting the user to perform a facial gesture. The method comprises recording a facial gesture performed by the user using a visual sensor. The method comprises comparing micro-movements of the user's facial features during the recorded facial gesture with a user-specific reference sample of micro-movements for a specific facial gesture. The method comprises authenticating the user based on a match between the micro-movements of the user's facial features during the recorded facial gesture and the user-specific reference sample of micro-movements for the specific facial gesture.