Content Modification Detection via Determination Image

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional digital watermark systems fail to detect modifications to content, such as text in images, leading to an increased burden for content providers in verifying authenticity.

Innovation Solution

A system that converts automatically generated content into determination information, generates a determination image from this information, and displays it alongside the content to facilitate modification detection, using hashing or encryption techniques to create a binary image that is not easily recognizable.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a predetermined ID watermark is embedded in content, then authentication and management of content become possible, but the ability to detect modifications to the content is lost

Engineering Contradiction:

Improvecontent authenticationVSAvoidmodification detection capability

Core Design Contradiction:

ReliabilityVSMeasurement precision

Solution Approach 1:

The content is segmented into multiple blocks, and a determination image is generated for each block based on its specific content. This allows modification detection at the block level rather than relying on a single overall watermark, enabling precise detection of local modifications while maintaining authentication.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

A determination image serves as an intermediary between the content and the authentication system. This determination image is generated from the content itself and used to verify both authenticity and integrity, bridging the gap between simple ID watermarking and comprehensive modification detection.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If conventional digital watermarking is used, then content can be authenticated, but the burden on content providers to verify modifications becomes excessive

Engineering Contradiction:

Improvecontent authenticationVSAvoidverification burden

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The content itself generates the determination image that will be used to verify its authenticity and integrity. This self-referential approach eliminates the need for external verification systems, reducing the burden on content providers while maintaining reliable authentication and modification detection.

Inventive Principle:

Principle #25Self-service

3Productivity

If text content is automatically generated and watermarked, then content creation efficiency is improved, but modification detection becomes impossible when text is replaced

Engineering Contradiction:

Improvecontent generation efficiencyVSAvoidtext modification detection

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

Different parts of the content (blocks) have different determination images generated specifically for each local region. This local quality approach ensures that modifications in any specific text portion can be detected, even in automatically generated content, without compromising the overall efficiency of content creation.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS20240354885A1Non-transitory computer-readable medium and system for content modification determination

Publication Date: 2024.10.24 SQUARE ENIX HLDG CO LTD

AI summary

One or more embodiments of the disclosure provide a non-transitory computer readable medium comprising a program that, when executed, causes a server to perform: displaying automatically generated content; converting at least a portion of the automatically generated content into information (hereinafter, this is referred to as “determination information”) used to determine whether the content has been modified; generating a determination image from the determination information; and displaying the determination image together with the content.