Video File Authorship Attribution via Segmented Fingerprinting

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The existing technologies face challenges in accurately attributing authorship of digital content, particularly video files, leading to misallocation of digital rights due to issues like 'freebooting' where content is re-hosted and modified, making it difficult to determine the original creator and allocate rights accordingly.

Innovation Solution

A method and system that utilize an authorship attribution database to examine video files, extract key segments, and determine authorship by comparing them to stored records, enabling accurate attribution and digital rights allocation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If content is re-hosted on different sites or by different authors, then content distribution and accessibility are improved, but authorship attribution accuracy deteriorates leading to misallocation of digital rights

Engineering Contradiction:
Improvecontent distributionVSAvoidauthorship attribution accuracy
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent segments video content into unique segments and creates fingerprints for each segment. When attributing authorship, the system divides the video file into segments and compares each segment against the database of segmented video files. This segmentation approach allows the system to handle re-hosted and modified content by identifying matching segments even when the overall video structure changes, thus maintaining attribution accuracy while enabling content distribution.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary mechanism - the segmented video file database with fingerprints - that acts as a mediator between content distribution and authorship verification. The database stores segmented versions of original videos with unique identifiers, allowing the system to verify authorship through intermediate comparison steps rather than direct whole-video matching, thereby resolving the contradiction between distribution flexibility and attribution precision.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If small revisions are made to video content, then content adaptability is improved, but authorship determination becomes difficult as second individuals may be incorrectly allocated digital rights

Engineering Contradiction:
Improvecontent modificationVSAvoidauthorship determination reliability
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The system segments video content into multiple unique segments and creates fingerprints for each segment. When a video with modifications is uploaded, the system segments it and compares each segment against the database. Even small revisions will result in some segments not matching the original, allowing the system to identify the extent of modification and maintain reliable authorship determination by detecting segment mismatches.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter of video comparison from whole-video matching to segment-level fingerprint matching. By transforming the comparison unit from the entire video to individual segments with unique identifiers, the system can detect parameter changes (modifications) at the segment level, thereby maintaining reliable authorship determination even when content adaptability through modification is improved.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If video files are examined using authorship attribution databases, then authorship attribution accuracy is improved, but processing time and computational resources increase

Engineering Contradiction:
Improveauthorship attribution accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments video files into smaller units and creates compact fingerprints for each segment. During authorship verification, the system processes segments independently rather than analyzing entire videos at once. This segmentation reduces the computational complexity and processing time while maintaining high attribution accuracy through cumulative segment matching.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system extracts only the essential identifying features (fingerprints) from video segments rather than processing and comparing entire video files. By taking out and storing only the critical fingerprint data in the database, the system enables rapid comparison and attribution without the time cost of processing complete videos, thus reducing processing time while preserving attribution precision.

Inventive Principle:
Principle #2Taking out (Extraction)

4Productivity

If key segments are extracted from video files, then authorship verification efficiency is improved, but information retention may be reduced

Engineering Contradiction:
Improveverification efficiencyVSAvoidcontent information retention
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The patent extracts key segments and their fingerprints from video files for storage in the database. This extraction focuses on capturing the essential identifying information needed for authorship verification while leaving the full video content intact for viewing and distribution. The extracted fingerprints serve as verification keys without compromising the original video's completeness.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system creates copies of key segment fingerprints rather than storing or removing actual video content. The fingerprints are copied and stored in the database for verification purposes, while the original complete video files remain unchanged and fully accessible. This copying approach enables efficient verification without losing any content information.

Inventive Principle:
Principle #26Copying

Data Source

PatentUS10289815B2Video file attribution
Publication Date: 2019.05.14 INTERNATIONAL BUSINESS MACHINE CORPORATION
  • US10289815B2 patent drawing
  • US10289815B2 patent drawing
  • US10289815B2 patent drawing

AI summary

Methods, computer program products, and systems are set forth herein and can include examining a video file using content of an authorship attribution database, the authorship attribution database including one or more authorship attribution video file record; and determining authorship attribution of the video file based on the examining. The authorship attribution database can include one or more authorship attribution video file record.