Audio Device Echo Filtering for Reliable Voice Control During Playback

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Multi-channel audio devices face challenges in reliably receiving voice operations during audio reproduction due to diffracted audio signals mixing with voice signals, leading to decreased recognition rates and increased processing demands.

Innovation Solution

An audio device that determines echo back parameters before audio reproduction to filter diffracted signals using voice operation reception microphones, improving voice recognition rates and reducing processing load during multi-channel audio reproduction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If real-time echo cancellation is implemented by analyzing correlation between audio signals and microphone signals for each speaker, then diffracted signals can be cancelled from voice operation reception, but processing capability and cost increase significantly

Engineering Contradiction:
Improvevoice recognition rateVSAvoidprocessing capability
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by pre-calculating and storing echo back parameters (transfer function characteristics) from speakers to microphones before voice operation reception. During actual voice recognition, these pre-computed parameters are directly applied to filter diffracted signals without real-time correlation analysis, significantly reducing processing requirements while maintaining voice recognition reliability

Inventive Principle:
Principle #10Preliminary action

2Adaptability or versatility

If voice operation reception function is added to multi-channel audio device, then usability is improved, but diffracted audio signals mix with voice signals causing recognition failure

Engineering Contradiction:
ImproveusabilityVSAvoidvoice recognition rate
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent introduces an intermediary filtering process using pre-computed echo back parameters that act as a mediator between the multi-channel audio reproduction system and voice operation reception. This filter selectively removes diffracted audio signals from the microphone input while preserving voice signals, enabling both audio reproduction and voice recognition to coexist reliably

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3605535B1Audio device
Publication Date: 2025.08.27 D & M HOLDINGS INC
  • EP3605535B1 patent drawingFigure 1
  • EP3605535B1 patent drawingFigure 2
  • EP3605535B1 patent drawingFigure 3

AI summary

[Problem) To provide an audio device capable of further reliably accepting voice control even during playback. [Solution] An audio device 2 comprises: a parameter determination unit 30 that determines an echo-back parameter for cancelling a wraparound signal that enters voice-control microphones 20-1, 20-2 as the audio from a speaker during playback; a filter processing unit 27 that performs filtering using the echo-back parameter for audio signals collected by the voice-control microphones 20-1, 20-2 during playback; a voice recognition unit 28 that performs voice recognition processing for a filtered audio signal; and an operation acceptance unit 29 that accepts voice control on the basis of the voice recognition results. The parameter determination unit 30 outputs a test signal from a speaker and collects the test signal through the voice-control microphones 20-1, 20-2, and determines the echo-back parameter on the basis of the delay time and the attenuation rate of the collected audio signal relative to the test signal.