Audio Device Echo Filtering for Reliable Voice Control During Playback
Find Innovative SolutionsGenerate 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
Engineering 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
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
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
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
Data Source
Figure 1
Figure 2
Figure 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.