Digital Assistant Trigger Detection Using Multi-Microphone Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Digital assistants face challenges in accurately detecting spoken triggers due to interfering audio signals and distinguishing relevant user inputs from irrelevant noise, leading to inefficient user experiences and incorrect session initiations.

Innovation Solution

The method involves sampling audio signals from multiple microphones to process multiple audio streams, determining if a spoken trigger is present, and initiating or foregoing a digital assistant session based on these signals, using a centralized device with the highest processing power to analyze and interpret the audio data accurately.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If audio signals are sampled from multiple microphones to improve detection accuracy, then measurement precision is improved, but device complexity increases

Engineering Contradiction:
Improvetrigger detection accuracyVSAvoidmicrophone array complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The audio detection process is segmented into distinct stages: initial trigger detection using a first microphone, followed by verification using a second microphone. This segmentation allows the system to maintain high detection accuracy while avoiding the continuous complexity of processing all microphones simultaneously, as each microphone is activated only when needed in the detection sequence.

Inventive Principle:
Principle #1Segmentation

2Speed

If audio signals are continuously monitored to improve response time, then speed is improved, but energy consumption increases

Engineering Contradiction:
Improvetrigger response timeVSAvoidpower consumption
Core Design Contradiction:
SpeedVSUse of energy by moving object

Solution Approach 1:

The system implements periodic action by continuously monitoring audio signals for trigger patterns and only activating the full multi-microphone verification process when a potential trigger is detected. During normal operation, the system uses lower-power monitoring, and only transitions to higher-power verification mode periodically when necessary, thus maintaining fast response times while reducing overall energy consumption compared to continuous full-power monitoring.

Inventive Principle:
Principle #19Periodic action

3Measurement precision

If multiple audio streams are processed to distinguish user input from noise, then measurement precision is improved, but loss of time increases

Engineering Contradiction:
Improveuser input distinction accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system applies preliminary action by pre-processing audio signals to identify characteristic trigger patterns before full verification. The first microphone continuously scans for potential triggers using simplified pattern recognition, and only when a match is found does the system activate the second microphone for comprehensive verification. This preliminary filtering reduces the total processing time required while maintaining high accuracy in distinguishing user input from noise.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3806091B1Detecting a trigger of a digital assistant
Publication Date: 2023.11.29 APPLE INC
  • EP3806091B1 patent drawingFigure 1
  • EP3806091B1 patent drawingFigure 2A
  • EP3806091B1 patent drawingFigure 2B

AI summary

Systems and processes for operating an intelligent automated assistant are provided. In accordance with one example, a method includes, at an electronic device with one or more processors, memory, and a plurality of microphones, sampling, at each of the plurality of microphones of the electronic device, an audio signal to obtain a plurality of audio signals; processing the plurality of audio signals to obtain a plurality of audio streams; and determining, based on the plurality of audio streams, whether any of the plurality of audio signals corresponds to a spoken trigger. The method further includes, in accordance with a determination that the plurality of audio signals corresponds to the spoken trigger, initiating a session of the digital assistant; and in accordance with a determination that the plurality of audio signals does not correspond to the spoken trigger, foregoing initiating a session of the digital assistant.