Asymmetric Microphone Array Loudspeaker Noise Attenuation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computer science techniques for audio localization using microphone arrays face challenges in distinguishing sound sources and determining their location, especially in videoconferencing environments where a loudspeaker acts as a constant noise source close to the microphones, interfering with the calculation of local sound sources.
Innovation Solution
An asymmetric microphone array geometry is employed, where microphones are arranged asymmetrically with respect to the loudspeaker to maximize distance and attenuate the loudspeaker's position in cross-correlation functions using time-delay windows and attenuation factors, allowing for robust audio localization despite the proximity of microphones to constant noise sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If microphones are placed close to the loudspeaker, then audio capture quality is improved, but the loudspeaker noise significantly impacts the ability to estimate local sound sources
Solution Approach 1:
The patent employs an asymmetric microphone array configuration where microphones are positioned at different distances and orientations relative to the loudspeaker. This asymmetric arrangement creates distinct acoustic paths that allow the system to differentiate between sound from the loudspeaker and sound from local participants, thereby reducing noise interference while maintaining localization precision.
Solution Approach 2:
The patent introduces time-delay windows and attenuation factors as intermediary processing elements. These parameters act as mediators that selectively suppress the loudspeaker noise component in the cross-correlation calculations while preserving the signal from local sound sources, enabling accurate audio localization despite the proximity of the loudspeaker to the microphones.
2Device complexity
If a symmetric microphone array is used, then the system is simpler to configure, but the loudspeaker position cannot be effectively differentiated from local sound sources
Solution Approach 1:
The patent transitions from a symmetric to an asymmetric microphone array configuration. This change introduces spatial differentiation in the microphone positions relative to the loudspeaker, creating unique acoustic signatures that enable the system to distinguish between loudspeaker noise and local sound sources, thereby improving location estimation precision.
Solution Approach 2:
The patent applies different processing characteristics to different microphone pairs based on their specific spatial relationship with the loudspeaker. By tailoring the time-delay window and attenuation factor to each microphone pair's local conditions, the system optimizes noise suppression while maintaining accurate localization for each spatial region.
3Object-affected harmful factors
If the loudspeaker is positioned far from microphones, then noise interference is reduced, but the system cannot effectively capture local audio in close-proximity environments
Solution Approach 1:
The patent implements dynamic adjustment of time-delay windows and attenuation factors based on the detected loudspeaker position and the specific microphone pair being processed. This dynamic processing allows the system to adapt to varying spatial relationships between the loudspeaker and microphones, maintaining effective noise suppression across different distances and configurations.
Solution Approach 2:
The patent utilizes variable time-delay window parameters and attenuation factors that can be adjusted based on the acoustic environment and loudspeaker position. By changing these parameters dynamically, the system maintains its ability to distinguish between loudspeaker noise and local sources regardless of the physical distance between the loudspeaker and microphones.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively extracts accurate audio localization information by minimizing the impact of constant noise sources, enhancing the ability to identify and isolate the active speaking participant in videoconferencing scenarios.
Implementation Method 1
sound data is received from a planar arrangement of microphones
Implementation Method 2
calculate the cross-correlation based on a time-delay window and an attenuation factor
Data Source
AI summary
In one implementation, an apparatus includes a camera, a loudspeaker, a plurality of microphones, and a controller. The apparatus may be a computer, a mobile device such as a smart phone, or a dedicated videoconference device. The plurality of microphones are configured to produce sound data. A first number of the plurality of microphones are arranged on a first side of the loudspeaker, and a second number of the plurality of microphones are arranged on a second side of the loudspeaker. Different quantities of the microphones may be included on each side of the loudspeaker. The controller is configured to calculate a cross-correlation of the plurality of microphones based on the sound data and attenuate the cross-correlation based on a time-delay window and an attenuation factor. A participant location is determined based on the attenuated cross correlation functions.