Voice-Activated Device Wake-Word Verification Against False Triggers

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Voice-activated intelligent devices are prone to false triggers due to unintentional detection of wake words or phrases from audio sources like televisions or radios, leading to privacy and security concerns.

Innovation Solution

Implement a method to confirm whether an audio input is intended to activate the device by analyzing the source of the input, using techniques such as fingerprint detection, inherent distortion analysis, or embedding ultrasonic signals to distinguish between human-generated speech and non-human audio signals.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If the device listens for wake words to enable quick response, then responsiveness is improved, but false positive activation increases

Engineering Contradiction:
Improveresponse speedVSAvoidfalse positive rate
Core Design Contradiction:
SpeedVSReliability

Solution Approach 1:

The system performs preliminary analysis of audio characteristics before triggering wake word detection. By pre-processing the audio signal to identify human speech patterns, the system prepares the detection mechanism in advance, allowing quick response only to genuine human speech while filtering out audio from televisions, radios, and other non-human sources.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary verification layer between audio detection and device activation. This intermediary mechanism analyzes audio fingerprints and speech characteristics to mediate between the wake word detector and the activation command, determining whether the detected wake word originates from human speech or non-human audio sources.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If audio fingerprinting is used to detect wake words in broadcasts, then false trigger detection is improved, but processing complexity increases

Engineering Contradiction:
Improvetrigger detection accuracyVSAvoidaudio processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system extracts only the essential audio fingerprint characteristics needed for wake word detection from the full audio signal. By taking out and analyzing only the relevant fingerprint features rather than processing the entire audio stream, the system achieves accurate detection of wake words in broadcasts while minimizing processing complexity and computational overhead.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP3633673B1False trigger correction for a voice-activated intelligent device
Publication Date: 2025.09.10 HARMAN INT IND INC
  • EP3633673B1 patent drawingFigure 1~5
  • EP3633673B1 patent drawingFigure 2A
  • EP3633673B1 patent drawingFigure 2B

AI summary

A method for confirming a trigger in an audio input detected by a voice-activated intelligent device that listens for and detects a trigger in the audio input, confirms whether the trigger in the audio input is intended to wake the device, and if confirmed, the device is instructed to activate. If the trigger cannot be confirmed, the device is instructed to ignore the trigger. The step of confirming whether the trigger is intended to activate the device may include determining whether the audio input is human generated speech, which may also include detecting a fingerprint in the audio input.