Face Authentication Material Detection Using Patterned IR Patches

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing material detection methods from images require intensive computing resources and often lack reliability, especially in complex scenarios where objects are covered by various materials, leading to potential spoofing attacks and reduced accuracy.

Innovation Solution

A method involving pattern image manipulation to generate partial images, which are processed by a data-driven model to extract material information, utilizing less training data and computational resources, thereby enhancing accuracy and preventing spoofing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If filtering techniques are used for material detection from images, then material information can be extracted, but intensive computing resources are required and reliability is reduced

Engineering Contradiction:
Improvematerial detection reliabilityVSAvoidcomputing resource consumption
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent segments the full image into multiple patches and processes each patch independently through the neural network. This segmentation approach reduces the computational burden on each processing unit while maintaining detection reliability through ensemble processing of multiple patches, directly addressing the contradiction between computing resource consumption and detection reliability

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent replaces traditional filtering techniques with a data-driven neural network model that has been trained to detect material properties. This substitution of the detection mechanism achieves higher reliability through learned features while the patch-based processing manages computational resources more efficiently

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

2Reliability

If traditional material detection methods are used, then material information can be obtained, but the methods lack reliability in complex scenarios where objects are covered by various materials

Engineering Contradiction:
Improveauthentication reliabilityVSAvoidhandling complex material scenarios
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

By dividing the image into patches, the system can handle complex scenarios where different materials are present in different regions. Each patch can be classified independently, allowing the system to adapt to varying material compositions across the object surface, thereby improving both reliability and adaptability

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The neural network model changes its detection parameters by being trained on diverse datasets representing various materials and scenarios. This enables the model to adapt to complex situations with multiple materials by adjusting its internal parameters based on learned patterns, improving reliability without sacrificing versatility

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If full images are used for material detection, then comprehensive information is available, but computing resources and training data requirements increase

Engineering Contradiction:
Improvematerial identification accuracyVSAvoidtraining data size
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent processes images as segmented patches rather than full images, which reduces the amount of data that needs to be processed and stored during training. This segmentation maintains measurement precision by allowing focused analysis of local material properties while significantly reducing the quantity of training data required

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent extracts only the necessary patches from full images for training and inference, removing unnecessary portions of the image data. This extraction approach maintains material identification accuracy by focusing on relevant regions while reducing the overall data volume that needs to be processed

Inventive Principle:
Principle #2Taking out (Extraction)

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 achieves accurate material identification and authentication with reduced computational expense, suitable for devices like smartphones, by using partial images and a data-driven model trained on less data, thus improving security and reliability.

Implementation Method 1

material information is derived from reflections of a pattern illumination

Methodology Applied
Scientific EffectReflection: Reflection

Data Source

PatentEP4479951B1Face authentication including material data extracted from image
Publication Date: 2026.04.08 TRINAMIX GMBH
  • EP4479951B1 patent drawingFigure 1
  • EP4479951B1 patent drawingFigure 2
  • EP4479951B1 patent drawingFigure 3

AI summary

The invention refers to a method for authorizing an object of a device to perform at least one operation on the device that requires authentication, comprising the steps: - receiving, from a user interface associated with a device, an unlock request for the device from a user; - in response to receiving the unlock request, triggering illumination of the object with patterned infrared illumination; - triggering to capture at least one image of the object using a camera located on the de- vice while the object is being illuminated with the patterned infrared illumination; - extracting material data from at least one image; and - allowing or declining the object to perform at least one operation on the device that requires authentication based on the material data.