Microphone Array Spatial Noise Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Far-field audio processing in smart home devices and similar systems faces challenges due to signal decay and interference from noise sources, leading to incorrect voice command recognition and confusion between multiple talkers.
Innovation Solution
The use of multiple microphones to spatially identify noise sources and filter them out, combined with beamforming techniques to enhance speech recognition by determining the location of talkers and configuring the microphone array to prioritize desired speech signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple microphones are used to capture far-field audio, then the coverage area is improved, but the signal-to-noise ratio deteriorates due to signal decay with distance
Solution Approach 1:
The audio signal processing is segmented into multiple independent channels, one for each microphone. Each channel processes the signal from its respective microphone independently, allowing for individual optimization of signal-to-noise ratio while maintaining broad spatial coverage through the combined array of microphones.
Solution Approach 2:
Signals from multiple microphones are merged through beamforming techniques that coherently combine the far-field signals while incoherently combining the noise. This merging process improves the overall signal-to-noise ratio by exploiting the spatial coherence of the desired signal across the microphone array.
2Area of stationary object
If multiple microphones are used to capture far-field audio, then the coverage area is improved, but the speech clarity deteriorates due to decreased direct-to-reverberant ratio
Solution Approach 1:
The beamforming process applies local quality enhancement by selectively emphasizing signals from specific spatial directions (where talkers are located) while suppressing signals from other directions (where reverberation and noise originate). This creates direction-dependent signal processing that improves speech clarity for far-field sources.
3Area of stationary object
If multiple microphones are used to capture far-field audio, then the coverage area is improved, but the interference from noise sources increases
Solution Approach 1:
The system converts the potentially harmful effect of having multiple noise sources in the environment into a benefit by using the spatial distribution of these noise sources. Through beamforming, the system exploits the spatial coherence properties to distinguish between correlated noise (from specific directions) and uncorrelated noise, thereby converting the presence of multiple noise sources into additional spatial information for interference suppression.
4Measurement precision
If beamforming is used to enhance speech recognition, then the speech clarity is improved, but the device complexity increases
Solution Approach 1:
The beamforming implementation uses parameter changes by adjusting the weights and phases of individual microphone signals based on the estimated positions of talkers. These parameter adjustments are dynamically computed and applied to enhance speech clarity while maintaining computational efficiency through adaptive algorithms that converge to optimal parameter settings.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach improves the accuracy of voice command recognition by reducing interference and distinguishing between multiple talkers, resulting in enhanced signal-to-noise ratios and more reliable speech recognition in noisy environments.
Implementation Method 1
Two or more microphones may be used to record sounds from the environment
Implementation Method 2
The received sounds processed to spatially identify noise sources. The identification of the noise sources may also be used to filter out the identified noise sources from the microphone signals
Data Source
AI summary
Information from microphone signals from a microphone array may be used to identify persistent sources, such as televisions, radios, washing machines, or other stationary sources. Values representative of broadside conditions for each pair of microphone signals are received from the microphone array. By monitoring broadside conditions for microphone pairs, a position of a sound source may be identified. If a sound source is frequently identified with a broadside of the same microphone pair, then that sound source may be identified as a persistent noise source. When a broadside of a pair of microphones is identified with a noise source, a beamformer may be configured to decrease contribution of that pair of microphones to an audio signal formed from the microphone array.