Inverted Client-Side Fingerprinting for Content Identification

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Fingerprinting systems face processing overload when generating and comparing fingerprints for large numbers of content items in real-time data streams, requiring significant hardware resources and costs, especially in large-scale systems monitoring millions of users.

Innovation Solution

Implementing inverted client-side fingerprinting and matching, where client devices generate and compare fingerprints locally, offloading the fingerprint generation and initial comparison to a central system, allowing a small number of fingerprinting systems to handle a large number of clients by transmitting generated fingerprints for initial identification and receiving reference fingerprints for ongoing matching.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a fingerprinting system generates and compares fingerprints for all content items in real-time data streams, then content identification accuracy is improved, but processing load and hardware resources required increase significantly

Engineering Contradiction:

Improvecontent identification accuracyVSAvoidhardware resources

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system divides the fingerprinting workload into two segments: client devices perform local fingerprint generation and initial comparison against cached reference fingerprints, while the central server handles fingerprint database management and receives results from clients. This segmentation distributes processing load across multiple devices, reducing the hardware resources required at any single point while maintaining identification accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Reference fingerprints are pre-generated and cached on client devices before real-time content identification is needed. This preliminary action allows clients to perform rapid local comparisons without requiring the server to generate fingerprints during real-time operations, significantly reducing the processing load on the fingerprinting system during actual content monitoring.

Inventive Principle:

Principle #10Preliminary action

2Productivity

If a fingerprinting system processes fingerprints for millions of users in real-time, then real-time content monitoring capability is improved, but processing load and system costs increase significantly

Engineering Contradiction:

Improvereal-time content monitoring capabilityVSAvoidprocessing load

Core Design Contradiction:

ProductivityVSLoss of energy

Solution Approach 1:

Client devices perform self-service by generating their own fingerprints locally and comparing them against cached reference fingerprints. This eliminates the need for the central server to process every fingerprint from every user, allowing the system to scale to millions of users without proportionally increasing server processing load or energy consumption.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

Reference fingerprints are pre-computed and cached on client devices, enabling rapid local comparisons during real-time content monitoring. This preliminary preparation allows the system to handle millions of concurrent users with minimal real-time processing requirements on the server side, significantly reducing processing load and energy consumption during operation.

Inventive Principle:

Principle #10Preliminary action

3Extent of automation

If fingerprint generation is performed centrally for all clients, then system control is improved, but processing load on the fingerprinting system increases significantly

Engineering Contradiction:

Improvesystem controlVSAvoidfingerprinting volume capacity

Core Design Contradiction:

Extent of automationVSProductivity

Solution Approach 1:

The system segments fingerprint generation responsibilities: clients generate their own fingerprints locally while the server maintains the reference fingerprint database and coordinates the identification process. This segmentation allows the system to maintain centralized control over the fingerprinting process while distributing the actual generation workload, thereby increasing overall fingerprinting volume capacity without overwhelming the server.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS9113202B1Inverted client-side fingerprinting and matching

Publication Date: 2015.08.18 GOOGLE LLC

AI summary

A technique for inverted client side fingerprinting and matching provides the benefits of disposable fingerprinting to identify multiple content streams from multiple clients without overloading a fingerprinting system. Rather than tasking a fingerprinting system with the generation and comparison of all fingerprints, the technique distributes some fingerprinting tasks to the clients receiving the content streams. As a result, the fingerprinting system is not bottlenecked by fingerprinting tasks. In one embodiment, the fingerprinting system can provide additional services to the clients.