Adaptive Noise-Cancelling Beamforming for Source Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio capture systems face challenges in optimal source separation, particularly in noisy environments, with conflicting desires for improved performance, reduced complexity, and adaptability to dynamic audio scenes, often resulting in undesired reverberation and noise interference.
Innovation Solution
An audio apparatus employing noise cancelling beamformers with adaptive filters and a hierarchical beamforming structure, including a noise cancelling beamformer and an adaptive beamformer, which adapt based on the presence of desired audio sources to enhance signal extraction and reduce noise and reverberation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If spatial beamforming is applied to separate audio sources, then source separation performance is improved, but device complexity increases
Solution Approach 1:
The audio processing system is segmented into multiple independent noise cancelling beamformers, each handling specific spatial regions or audio sources. This segmentation allows the complex beamforming task to be divided into manageable modules, improving source separation while controlling overall system complexity through modular architecture
Solution Approach 2:
The beamforming system employs dynamic adaptation where beamformers are selectively activated or deactivated based on detected audio source presence. The adaptation circuit dynamically adjusts the operational state of different beamformers, optimizing performance for current acoustic scenarios while reducing computational load when fewer sources are present
2Object-affected harmful factors
If adaptive filtering is applied to cancel noise, then noise reduction is improved, but device complexity increases
Solution Approach 1:
The noise cancellation function is segmented across multiple parallel beamformers, each with its own adaptive filter. This distribution of filtering tasks allows each individual filter to operate on reduced complexity while collectively achieving superior noise reduction through the combined output of multiple specialized filters
Solution Approach 2:
The system uses multiple copies of the adaptive filter structure across different beamformers. Each beamformer contains a complete adaptive filtering module that processes its specific input signals, creating redundant but specialized filtering paths that collectively provide robust noise cancellation without requiring a single complex monolithic filter
3Reliability
If multiple beamformers are used to improve audio capture, then audio quality is improved, but device complexity increases
Solution Approach 1:
Multiple beamformer outputs are merged by the adaptation circuit through selective combination based on detected source presence. The system combines outputs from active beamformers while excluding those corresponding to absent sources, achieving improved audio capture quality through intelligent merging rather than simple summation, thereby managing system complexity through selective integration
4Adaptability or versatility
If adaptation is continuously applied to track audio sources, then adaptability is improved, but device complexity increases
Solution Approach 1:
The adaptation mechanism operates dynamically by continuously monitoring audio source presence and adjusting beamformer activation states accordingly. The adaptation circuit implements dynamic control where filtering and beamforming parameters are adjusted in real-time based on detected acoustic conditions, improving adaptability while managing complexity through event-driven rather than continuous adjustment
Solution Approach 2:
The system employs feedback mechanisms where the detected presence or absence of audio sources feeds back to the adaptation circuit, which then adjusts the operational state of beamformers. This closed-loop feedback control enables automatic adaptation to changing audio scenes without requiring complex manual configuration or continuous optimization algorithms
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An audio apparatus comprises a plurality of noise cancelling beamformers (105) each comprising: a beamform circuit (201) generating a beamformed signal capturing audio in a beam and a noise reference signal capturing audio outside the beam, an adaptive filter (203) filtering the noise reference signal to generate a filtered noise signal, and a compensation circuit (205) generating the beamformed audio signal by removing the filtered noise signal from the beamformed signal. An adaptive beamformer (107) generates an output audio signal by applying a beamform operation to the beamformed audio signals. An audio detector (115) detects a presence of a desired audio source in the plurality of audio signals and an adaptation circuit (109) adapts the adaptive beamformer (107) and the adaptive filters (203). The adaptation circuit (109) adapts the adaptive beamformer (107) and does not adapt the adaptive filters (203) when the presence of the desired audio source is detected. Improved audio/speech capture can often be achieved for strong interference and noise.