Constrained Beamforming for Audio Capture in Reverberant Environments

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional beamforming systems for audio capture, particularly in noisy and reverberant environments, face challenges in distinguishing between echoes and diffuse background noise, leading to speech distortion and slower convergence of adaptive beams, especially when the speaker is outside the reverberation radius.

Innovation Solution

The system employs a combination of adaptive and constrained beamformers, where the adaptive beamformer is supplemented by constrained beamformers that adapt only when specific criteria are met, and the adaptation rate is controlled to ensure robust and accurate beam formation, even for audio sources outside the reverberation radius.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If adaptive beamforming is used to extract desired speech sources, then speech capture performance is improved, but convergence speed decreases in noisy and reverberant environments

Engineering Contradiction:
Improvespeech capture performanceVSAvoidconvergence speed
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system segments the beamforming process into multiple independent beams (first beam, second beam, third beam) with different characteristics. Each beam processes audio signals independently, allowing parallel operation that accelerates overall convergence while maintaining reliable speech capture through diverse beam strategies

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system dynamically switches between different beam configurations based on environmental conditions. The beam selection and parameter adjustment adapt to noisy and reverberant environments, optimizing convergence speed while maintaining speech capture reliability through real-time dynamic adaptation

Inventive Principle:
Principle #15Dynamics

2Reliability

If beamforming parameters are adapted to focus on speech sources, then speech extraction is improved, but speech distortion occurs due to echo and diffuse background noise

Engineering Contradiction:
Improvespeech extraction accuracyVSAvoidspeech distortion
Core Design Contradiction:
ReliabilityVSObject-generated harmful factors

Solution Approach 1:

The system segments the speech extraction process into multiple specialized beams. The first beam focuses on speech sources, the second beam targets echo suppression, and the third beam handles diffuse background noise. This segmentation allows each beam to specialize in reducing specific types of distortion while maintaining extraction accuracy

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system converts harmful echoes and diffuse noise into useful information by creating dedicated beams to detect and characterize these干扰 sources. The second and third beams specifically target echo and diffuse noise patterns, transforming these harmful factors into detectable signals that can be selectively suppressed while preserving speech quality

Inventive Principle:
Principle #22Blessing in disguise (Convert harm into benefit)

3Adaptability or versatility

If multiple beamformers are used to handle different audio sources, then adaptability to diverse environments is improved, but device complexity increases

Engineering Contradiction:
Improveenvironmental adaptabilityVSAvoidsystem complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system segments beamforming functionality into three specialized beam modules that can be independently configured and processed. Each beam handles specific audio source types, allowing the system to achieve high environmental adaptability through modular segmentation rather than requiring a single complex adaptive beamformer

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system achieves environmental adaptability by changing beam parameters (direction, width, depth) rather than changing the fundamental beamforming structure. This parameter-based adaptation maintains simplicity by avoiding additional hardware or complex architectural changes while still providing versatility across diverse audio environments

Inventive Principle:
Principle #35Parameter changes

4Area of stationary object

If beamforming focuses on distant speakers outside the reverberation radius, then coverage area is improved, but noise suppression performance deteriorates

Engineering Contradiction:
Improveaudio source coverage areaVSAvoidnoise suppression performance
Core Design Contradiction:
Area of stationary objectVSReliability

Solution Approach 1:

The system segments noise suppression into specialized beams including a third beam configured specifically for diffuse background noise. This segmentation allows distant speakers to be captured with extended coverage while dedicated beams handle the increased noise challenge, maintaining suppression performance despite the larger coverage area

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system applies different beam characteristics to different spatial regions and noise types. The third beam is specifically configured with parameters optimized for diffuse background noise suppression, providing local quality enhancement in noise-prone areas while maintaining broad coverage for distant speakers through other beam configurations

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3566461B1Method and apparatus for audio capture using beamforming
Publication Date: 2021.11.24 KONINKLIJKE PHILIPS NV
  • EP3566461B1 patent drawingFigure 1
  • EP3566461B1 patent drawingFigure 2
  • EP3566461B1 patent drawingFigure 3

AI summary

An apparatus for capturing audio comprises a first beamformer (305) coupled to a microphone array (301) and arranged to generate a first beamformed audio output. A plurality of constrained beamformers (309, 311) each generates a constrained beamformed audio output. A first adapter (307) adapts beamform parameters of the first beamformer (305) and a second adapter (313) adapts constrained beamform parameters for the plurality of constrained beamformers (309, 311). A difference processor (317) determines a difference measure for the constrained beamformers (309, 311) where the difference measure is indicative of the difference between beams formed by the first beamformer (305) and the constrained beamformers (309, 311). The second adapter (313) is arranged to adapt constrained beamform parameters with the constraint that beamform parameters are adapted only for constrained beamformers of the plurality of constrained beamformers (309, 311) for which a difference measure has been determined that meets a similarity criterion.