Keyword Detection Apparatus Preventing False Triggers

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional voice-controlled apparatuses falsely detect keywords or phonemes similar to intended keywords, leading to unintended keyword detection.

Innovation Solution

A keyword detection apparatus comprising a keyword detection unit, a voice detection unit, a delay unit, and an in-sentence keyword excluding unit, which generates a delayed voice section detection result to differentiate between intended and unintended keyword utterances, preventing false detections.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If keyword detection is performed without voice section detection, then keyword detection speed is improved, but false detection rate increases

Engineering Contradiction:
Improvekeyword detection speedVSAvoidfalse detection rate
Core Design Contradiction:
SpeedVSReliability

Solution Approach 1:

The voice detection unit performs preliminary detection of voice sections before the keyword detection unit processes the audio signal. By pre-identifying which time segments contain actual speech, the system can later focus keyword detection only on those segments, preventing false detections from non-voice audio while maintaining efficient processing speed.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The voice section detection result acts as an intermediary filter between the raw audio input and the keyword detection process. This intermediary component separates relevant voice-containing segments from irrelevant silent or noise segments, allowing the keyword detection unit to operate faster on reduced data without sacrificing accuracy.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If voice section detection is added to differentiate intended and unintended keyword utterances, then false detection accuracy is improved, but device complexity increases

Engineering Contradiction:
Improvefalse detection accuracyVSAvoiddevice complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The audio processing system is segmented into distinct functional modules: a voice detection unit that identifies voice sections, and a keyword detection unit that processes only those voice sections for keyword detection. This segmentation allows each unit to specialize in specific tasks, improving overall precision while keeping the architecture modular and manageable despite the increased complexity.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3852099B1Keyword detection apparatus, keyword detection method, and program
Publication Date: 2024.01.24 NIPPON TELEGRAPH & TELEPHONE CORP
  • EP3852099B1 patent drawingFigure 1
  • EP3852099B1 patent drawingFigure 2~3
  • EP3852099B1 patent drawingFigure 4A~4B

AI summary

An object is to prevent a keyword from being falsely detected from an utterance that is not intended for keyword detection. A keyword detection unit 11 generates a keyword detection result indicating a result of detecting an utterance of a predetermined keyword from an input voice. A voice detection unit 12 generates a voice section detection result indicating a result of detecting a voice section from the input voice. A delay unit 13 gives a delay that is at least longer than an utterance time of the keyword, to the voice section detection result. An in-sentence keyword excluding unit 14 updates the keyword detection result to a result indicating that the keyword has not been detected when the keyword detection results indicates that the keyword has been detected and the voice section detection result indicates that a voice section has been detected.