Microphone Wind Noise Detection Using Complex Coherence Phase
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise detection algorithms in electronic devices, particularly those used in vehicles, struggle to accurately differentiate between wind noise and other noise sources, leading to false wind noise detection and ineffective speech detection due to assumptions about phase differences between microphones, which can be caused by unmatched microphones or acoustic reflections.
Innovation Solution
The system employs complex coherence analysis to determine the phase of audio signals from multiple microphones, using power spectral densities and phase distribution to estimate wind and speech presence, and applies noise suppression techniques based on these determinations to differentiate between coherent speech and incoherent wind noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If phase difference assumptions are used to differentiate wind noise from speech, then wind noise detection is simplified, but false positives increase due to unmatched microphones or acoustic reflections
Solution Approach 1:
The patent changes the detection parameter from phase difference to complex coherence. Complex coherence is calculated as the normalized cross-spectral density between microphone signals, providing a more robust metric that accounts for amplitude and phase relationships without being sensitive to the limitations of simple phase difference assumptions. This parameter change resolves the contradiction by maintaining algorithmic simplicity while significantly improving detection accuracy in the presence of unmatched microphones and acoustic reflections.
2Measurement precision
If complex coherence analysis is used to improve wind and speech detection accuracy, then false positives are reduced, but computational complexity increases
Solution Approach 1:
The patent segments the audio signal processing into distinct frequency bins using Fast Fourier Transform (FFT). Complex coherence is calculated independently for each frequency bin, allowing the system to process different frequency components separately. This segmentation reduces computational complexity by enabling efficient parallel processing and avoiding the need to analyze the entire spectrum as a single complex problem, while still achieving accurate wind and speech differentiation across all frequencies.
Solution Approach 2:
The patent applies complex coherence analysis selectively rather than uniformly across all signal processing operations. The system calculates complex coherence primarily for wind noise detection purposes, while using simpler methods for other audio processing tasks. This partial application of the complex methodology reduces overall computational burden while maintaining high accuracy where it is most needed - in distinguishing wind noise from speech.
Data Source
AI summary
Techniques for improving microphone noise detection and suppression are provided. A method for detecting wind noise corresponding to microphone signals may include determining a phase of a complex coherence of the microphone signals. The phase may be used to determine a presence of wind near the microphones. A derivative of the phase may be used to determine a presence of speech near the microphones. Further, a method for suppressing noise caused by the wind may include determining a gain based on a cross power spectrum of the microphone signals and applying the gain to the microphone signals. The method for suppressing the noise may also include attenuating the microphone signals using a post filter. Based on detection of the wind, microphones which are more exposed to the wind may not be used to process the speech whereas microphones less exposed to the wind may be used to process the speech.