Image-Based Data Encoding for Confidential Consistency Checks

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods struggle to securely compare and assess the consistency of private data sets while preserving the confidentiality of the underlying data.

Innovation Solution

A method involving encoding data values into images and using an image-classification machine learning model to generate numerical output values, allowing secure comparison of data sets without revealing the actual data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If data values are directly shared between computing devices for comparison, then data consistency can be assessed, but data confidentiality is compromised

Engineering Contradiction:

Improvedata consistency assessmentVSAvoiddata confidentiality

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The patent introduces an intermediary transformation layer where data values are converted to images and processed through a machine learning model to generate numerical outputs. These numerical outputs serve as intermediaries that preserve the relative relationships between data values while masking the actual data content, enabling consistency assessment without data exposure

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent creates a transformed representation (numerical output) that copies the essential comparison properties of the original data while lacking the data's actual meaning. The numerical output preserves ordering and relative magnitude relationships needed for consistency checking but discards the specific data values themselves

Inventive Principle:

Principle #26Copying

2Loss of information

If data values are encoded into images and processed through machine learning models, then data confidentiality is preserved, but the complexity of the system increases

Engineering Contradiction:

Improvedata confidentialityVSAvoidsystem complexity

Core Design Contradiction:

Loss of informationVSDevice complexity

Solution Approach 1:

The patent employs a universal approach by using image classification models that can process various types of data representations. The same image processing pipeline works for different data types and comparison scenarios, reducing the need for multiple specialized systems and actually simplifying the overall architecture despite the transformation steps

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS20250371851A1Data encoding using images and machine learning models

Publication Date: 2025.12.04 SAP SE

AI summary

A method includes obtaining a plurality of first data values; creating a first image comprising first pixels, wherein the number of first pixels is equal to the number of first data values and wherein each first data value is assigned to a respective first pixel; providing the first image as input to an image-classification machine learning model to obtain a first numerical output value; obtaining a plurality of second data values; creating a second image comprising a plurality of second pixels equal to the number of second data values and wherein each of the plurality of second data values is assigned to a respective second pixel providing the second image as input to the model to obtain a second numerical output value; evaluating the first numerical output value and the second numerical output value to determine whether the first data values are consistent with the second data values.