Incoherent Noise Estimation Across Multiple Microphones
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately estimate and mitigate incoherent noise, such as wind noise, in sound recordings due to varying noise levels across different microphones in devices, which are affected by physical characteristics like shadowing and relative microphone positions.
Innovation Solution
A method and apparatus that utilize multiple microphones to make a primary noise estimation, determine direction and diffuseness parameters, and adjust microphone signal levels based on predicted level differences to accurately estimate noise levels, accounting for physical characteristics and noise sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing noise estimation methods are used, then the process is simple, but the noise level estimation accuracy is poor due to varying noise levels across microphones affected by shadowing and relative positions
Solution Approach 1:
The noise estimation process is segmented into multiple independent steps: primary noise estimation, direction parameter determination, diffuseness parameter determination, predicted level difference estimation, and final noise level calculation. Each step processes specific parameters separately, allowing for improved accuracy while maintaining manageable complexity through modular processing.
Solution Approach 2:
The method performs preliminary actions by determining direction parameters and diffuseness parameters before final noise level estimation. These preliminary parameters characterize the acoustic environment and are used to adjust the predicted level differences, improving the accuracy of the final noise measurement by accounting for shadowing and microphone position effects in advance.
2Measurement precision
If multiple microphones are used with direction and diffuseness parameter determination, then noise level estimation accuracy improves, but the computational complexity increases
Solution Approach 1:
The system dynamically adjusts the noise estimation process based on the primary noise estimation results. Direction parameters are determined when the primary estimation indicates significant noise, while diffuseness parameters are determined based on additional thresholds. This dynamic approach optimizes computational resources by only performing complex calculations when necessary, rather than always processing at maximum complexity.
Solution Approach 2:
The method uses feedback mechanisms where the primary noise estimation results inform subsequent processing steps. The determined direction and diffuseness parameters feed into the predicted level difference calculation, which then refines the final noise level estimation. This feedback loop allows the system to adapt its processing complexity based on the actual noise conditions detected.
Data Source
AI summary
To implement some methods of attenuating noise it can be useful to obtain an accurate estimate of the noise in the microphone signals. Examples of the disclosure provide a method of accurately estimating incoherent noise levels. In examples of the disclosure a primary estimation of a noise amount is made. The primary estimation of a noise amount is used to determine a direction parameter and a diffuseness parameter. The determined direction parameter and the estimated diffuseness parameter are used to estimate a predicted level difference between a first microphone signal and a second microphone signal. A level difference between the first microphone signal and the second microphone signal is determined and a noise level is estimated using at least the determined level difference and the predicted level difference.