Wind Noise Detection in In-Car Communication Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing in-car communication systems face challenges in effectively reducing wind noise, which is non-stationary and highly variable, leading to poor speech quality in vehicles due to its differing spectral characteristics and independence from speech signals.
Innovation Solution
A wind noise module is integrated into the ICC system to detect wind noise using log-power ratio covariance and spectral features, applying compensation factors to attenuate or filter the microphone signals, particularly in low-frequency bands, and adjusting processing to minimize wind noise impact.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If conventional noise reduction methods are used, then stationary or slowly varying background noise can be reduced, but wind noise cannot be effectively reduced due to its non-stationary and highly variable characteristics
Solution Approach 1:
The system dynamically adapts its noise reduction strategy by detecting wind noise presence through spectral analysis and power level comparisons. When wind noise is detected, the system switches from conventional stationary noise reduction to wind-specific processing, including adjusting beamforming weights and applying frequency-dependent attenuation. This dynamic adaptation allows the system to handle the non-stationary nature of wind noise effectively.
Solution Approach 2:
The system changes processing parameters based on detected wind noise conditions. It modifies the noise reduction filter characteristics, adjusts the beamforming vector weights, and changes the frequency bands targeted for attenuation. By dynamically changing these parameters according to the non-stationary wind noise profile, the system maintains effective noise reduction despite the varying noise characteristics.
2Object-affected harmful factors
If wind noise reduction processing is applied to all frequency bands, then wind noise is reduced, but speech quality may be degraded due to over-attenuation of useful signals
Solution Approach 1:
The system applies different attenuation characteristics to different frequency bands rather than uniform processing. It identifies frequency regions where wind noise dominates and applies strong attenuation there, while preserving or lightly processing frequency bands where speech content is present. This localized, frequency-selective approach reduces wind noise without degrading speech quality.
Solution Approach 2:
The system continuously monitors the processed signal and uses feedback to adjust processing parameters. By analyzing the output after noise reduction, it can detect if speech quality is being degraded and automatically adjust the attenuation levels or filter characteristics to preserve useful speech information while maintaining wind noise reduction.
3Adaptability or versatility
If the system processes audio signals from multiple acoustic zones simultaneously, then comprehensive communication coverage is achieved, but system complexity increases due to multiple signal processing paths
Solution Approach 1:
The system divides the passenger compartment into distinct acoustic zones (front, rear, left, right) and processes signals from each zone independently through dedicated microphone arrays and processing paths. This segmentation allows the complex task of multi-zone audio processing to be broken down into manageable, modular units, each handled by specialized processing logic that can be optimized independently.
Solution Approach 2:
The beamforming and noise reduction processing modules are designed to serve multiple acoustic zones simultaneously. The same core processing algorithms and hardware resources can be applied across different zones, with only the microphone array selections and spatial parameters varying. This multi-functional design reduces overall system complexity compared to having completely separate processing systems for each zone.
Data Source
Figure 1
Figure 2~3
AI summary
An in-car communication (ICC) system has multiple acoustic zones having varying acoustic environments. At least one input microphone within at least one acoustic zone develops a corresponding microphone signal from one or more system users. At least one loudspeaker within at least one acoustic zone provides acoustic audio to the system users. A wind noise module makes a determination of when wind noise is present in the microphone signal and modifies the microphone signal based on the determination.