Coherence-Based Noise Estimation for Adaptive Vehicle Audio Gain
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Variable acoustic noise in moving vehicles degrades the quality of music or speech by masking soft sounds and altering the fidelity or intelligibility, requiring frequent adjustments in audio system volume.
Innovation Solution
A method and system for dynamic sound adjustment that estimates the power spectral density of noise using a cross-spectral density matrix, iteratively modifying the frequency domain representation to separate noise from audio signals, and generating a control signal to adjust acoustic transducer gains based on noise levels, thereby maintaining a consistent audio experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the volume of the audio system is increased to compensate for noise, then the audio quality is improved in noisy conditions, but the volume becomes too high when noise decreases, requiring manual adjustment
Solution Approach 1:
The system continuously monitors the acoustic environment using a microphone to detect noise levels and automatically adjusts the audio system volume in real-time. The controller receives feedback about the acoustic conditions and modifies the gain of acoustic transducers accordingly, eliminating the need for manual volume adjustments and maintaining consistent audio quality across varying noise conditions
Solution Approach 2:
The audio system performs self-adjustment by automatically detecting noise levels and regulating its own output volume. The system uses its own acoustic transducers and microphone to monitor the acoustic environment and autonomously modifies the audio output without requiring user intervention, making the system self-regulating in response to changing acoustic conditions
2Measurement precision
If traditional noise estimation methods are used, then the system can detect noise levels, but computational resources are wasted by determining time waveform of noise signal first before computing frequency-specific information
Solution Approach 1:
Instead of following the traditional approach of first determining the time waveform of the noise signal and then computing frequency-specific information, the system inverts the process by directly computing frequency-specific noise information from the microphone signal using spectral analysis. This is achieved by deriving the power spectral density of noise directly without wasting computing resources in determining a time waveform of the noise signal first
Solution Approach 2:
The system extracts only the necessary frequency-specific noise information directly from the microphone signal through spectral analysis. By using the cross-spectral density matrix and coherence values, the system extracts the power spectral density of noise components without the intermediate step of determining the complete time waveform, obtaining only the relevant frequency-domain characteristics needed for audio adjustment
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The technology described herein can be embodied in a method for estimating a power spectral density of noise, the method including receiving an input signal representing audio captured using a microphone. The input signal includes a first portion that represents acoustic outputs from two or more audio sources, and a second portion that represents a noise component. The method also includes iteratively modifying a frequency domain representation of the input signal, such that the modified frequency domain representation represents a portion of the input signal in which effects due to the first portion are substantially reduced. The method further includes determining, from the modified frequency domain representation, an estimate of a power spectral density of the noise, and generating a control signal configured to adjust one or more gains of an acoustic transducer. The control signal is generated based on the estimate of the power spectral density of the noise.