Speech Context Authenticator Reducing False Positives

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing speech authentication systems face challenges in minimizing false positives and negatives due to variations in speech patterns such as accent, intonation, pitch, and emotional changes, which can be mimicked or altered.

Innovation Solution

A computer-implemented method and system that authenticates speech by receiving electronic voice communication and incorporating biometric context information, such as emotional and locational data, using a speech with context model to verify the identity of the speaker, thereby providing a two-factor authentication.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If speech authentication uses only speech patterns (accent, intonation, pitch), then the authentication process is simple, but false positives increase due to speech copying and variation

Engineering Contradiction:

Improveauthentication process complexityVSAvoidauthentication accuracy

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The patent combines speech pattern analysis with context analysis (biometric information, emotional state, locational data) into a unified authentication system. The speech authentication module and context authentication module work together to evaluate both the speech characteristics and the contextual factors, merging multiple authentication dimensions to improve reliability while maintaining reasonable complexity

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent transitions from one-dimensional speech pattern matching to multi-dimensional authentication by adding context as a new dimension. This includes biometric information, emotional state, and locational context, creating a holistic authentication framework that evaluates speech within its contextual framework, thereby reducing false positives and negatives

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Productivity

If speech authentication considers only speech characteristics, then processing is fast, but false negatives increase due to natural speech variation

Engineering Contradiction:

Improveauthentication processing speedVSAvoidauthentication accuracy

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system performs preliminary analysis of context information (biometric data, emotional state, location) alongside speech patterns. By evaluating multiple authentication factors simultaneously rather than sequentially, the system maintains processing speed while improving accuracy through comprehensive assessment

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The authentication system incorporates feedback mechanisms where the context information reinforces or challenges the speech pattern analysis. The speech-to-text conversion, emotional state detection, and location verification provide feedback loops that help distinguish genuine speech variations from spoofing attempts, reducing false negatives

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS10672402B2Speech with context authenticator

Publication Date: 2020.06.02 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A computer-implemented method for authenticating speech includes executing on a computer processor the step of receiving speech and context for the speech of a speaker, wherein the speech is received as electronic voice communication, wherein the context for the speech includes biometric information measured during the received speech. The speaker is authenticated according the received speech and the context by a speech with context model. A result of the authentication is returned.