Dual-Microphone Reverberation Mitigation Using Coherence Gain
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio processing systems struggle to effectively suppress reverberant components in audio signals recorded by external microphones, particularly in enclosed spaces, which hinders speech intelligibility and degrades overall speech quality, especially in environments with reflective surfaces.
Innovation Solution
A dual microphone signal processing system that converts time-domain signals to the time-frequency domain, applies binary and sigmoid gain functions based on energy ratios and inter-microphone coherence, and then transforms back to the time-domain to reduce reverberation components, without requiring prior knowledge of the clean signal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single microphone is used to capture audio signals, then the device complexity is low, but the speech intelligibility deteriorates due to reverberation
Solution Approach 1:
The patent divides the audio capture task across multiple microphones (at least two microphones) to spatially separate direct sound from reverberant components. This segmentation allows the system to capture different acoustic perspectives simultaneously, enabling subsequent separation and suppression of reverberation through comparative analysis of the captured signals.
Solution Approach 2:
The patent transitions from temporal signal processing alone to a spatio-temporal approach by incorporating spatial information from multiple microphone positions. This adds a spatial dimension to the signal processing, enabling the system to distinguish between direct and reverberant components based on their different spatial characteristics and time delays.
2Reliability
If prior knowledge of the clean signal is required for reverberation suppression, then the speech quality improves, but the adaptability deteriorates in realistic scenarios
Solution Approach 1:
The patent enables the system to process reverberant signals without requiring external clean signal references. The method uses only the signals captured by the multiple microphones to compute the binary mask and suppress reverberation, making the system self-sufficient and adaptable to any acoustic environment where the microphone array can capture the reverberant speech.
Solution Approach 2:
The patent transforms the signal into the time-frequency domain and applies frequency-specific energy ratio comparisons to dynamically adapt the processing parameters. This allows the system to adjust its reverberation suppression strategy based on the instantaneous spectral characteristics of the captured signal, improving both quality and adaptability.
3Device complexity
If linear prediction analysis is used for binary mask estimation, then the computational complexity is low, but the measurement precision deteriorates in noisy environments
Solution Approach 1:
The patent segments the frequency spectrum into multiple bins and computes energy ratios independently for each bin. This frequency-domain segmentation allows the system to process different frequency regions separately, improving the accuracy of binary mask estimation by focusing energy comparisons at specific frequencies where speech components are more distinguishable from noise and reverberation.
Data Source
AI summary
A dual microphone signal processing arrangement for reducing reverberation is described. Time domain microphone signals are developed from a pair of sensing microphones. These are converted to the time-frequency domain to produce complex value spectra signals. A binary gain function applies frequency-specific energy ratios between the spectra signals to produce transformed spectra signals. A sigmoid gain function based on an inter-microphone coherence value between the transformed spectra signals is applied to the transformed spectra signals to produce coherence adapted spectra signals. And an inverse time-frequency transformation is applied to the coherence adjusted spectra signals to produce time-domain reverberation-compensated microphone signals with reduced reverberation components.