Microphone Array Muting With Obstacle-Aware Attenuation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional audio systems in video conferencing and telepresence systems cannot independently mute specific sound sources, leading to leakage of muted speech signals and distortion in unmuted zones, especially in complex acoustic environments with obstacles.
Innovation Solution
The implementation of algorithms that identify the strongest microphone signal, estimate the location of the active speaker, determine if it's affected by obstacles, and apply modified attenuation to microphones in unmuted zones to prevent leakage of muted speech, using relative signal level matrices for calibration and image processing to track obstacles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional audio systems mute all microphones in a zone, then speech from muted zones can be suppressed, but speech leakage still occurs and sound quality in unmuted zones deteriorates
Solution Approach 1:
The system segments the audio signal processing by identifying individual sound sources within zones and applying muting selectively to specific microphones rather than all microphones in a zone. The processor divides the audio field into multiple zones and further segments each zone into sound sources based on signal strength and spatial information, enabling precise muting control at the sound source level.
Solution Approach 2:
The system applies different processing qualities to different spatial locations and sound sources. Microphones capturing speech from muted zones receive attenuation while microphones in unmuted zones maintain full signal quality. The processor dynamically adjusts attenuation levels based on the specific microphone's spatial relationship to muted and unmuted sound sources, creating local quality differentiation across the audio field.
2Reliability
If signal processing is applied to remove sound from muted zones, then speech leakage is reduced, but computation complexity increases
Solution Approach 1:
The system performs preliminary calibration by establishing relative signal level matrices (RSLM) and relative signal level matrices for high frequency (RSLM-H) during setup periods when no speech is present. These pre-computed reference matrices capture the acoustic characteristics of the environment including obstacle positions, enabling the processor to quickly determine attenuation levels during active speech without performing complex real-time acoustic modeling.
Solution Approach 2:
The system continuously monitors microphone signal strengths and compares them against the pre-stored RSLM and RSLM-H matrices to dynamically adjust attenuation levels. The processor uses feedback from actual signal measurements combined with the reference matrices to determine optimal attenuation for each microphone, balancing speech suppression effectiveness with computational efficiency.
3Measurement precision
If obstacles are present in the acoustic environment, then sound propagation is affected, but accurate sound source localization becomes more difficult
Solution Approach 1:
The system replaces complex acoustic modeling and simulation approaches with a signal-processing-based localization method. Instead of modeling sound wave propagation through obstacles using acoustic physics, the system uses signal strength comparisons across multiple microphones combined with pre-stored reference matrices to estimate sound source locations and determine obstacle effects, achieving accurate localization without requiring complex acoustic simulations.
Data Source
AI summary
A processing system can include tracking microphone array(s), audio-tracking circuitry configured to detect a location of audio sources from audio signals from the array(s), and processing circuitry. The processing circuitry can be configured to: identify a first microphone that has a strongest signal strength; estimate a location of an active speaker based on at least an output of the audio-tracking circuitry; determine whether a second microphone for the active speaker is affected by an acoustic obstacle based on the location of the active speaker and a location of the first microphone that has the strongest signal strength; estimate attenuation for microphones based on a comparison of actual signal strengths of the microphones with estimated signal strengths of the microphones that are estimated based on microphone signals of the second microphone for the active speaker; and modify the attenuation based on an estimated location of the acoustic obstacle.