Bilateral Microphone Array for Speech Intelligibility

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional conversation assistance systems, including head-mounted and off-head microphone arrays, face challenges in improving speech intelligibility due to limitations in directivity, self-noise amplification, and lack of spatial cues, especially in noisy environments.

Innovation Solution

A bi-lateral microphone array system with a processor that creates separate left-ear and right-ear audio signals using multiple microphones arranged externally, incorporating polar specifications based on head-related transfer functions and adjustable directivity to enhance speech intelligibility and reduce noise amplification.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If array directivity is increased to improve speech intelligibility, then talker-to-noise ratio improves, but self-noise amplification increases

Engineering Contradiction:
Improvespeech intelligibilityVSAvoidself-noise
Core Design Contradiction:
Measurement precisionVSObject-generated harmful factors

Solution Approach 1:

The patent applies white noise gain (WNG) constraints as a parameter control mechanism to limit the amplification of uncorrelated noise while maintaining array directivity for speech enhancement. By constraining the WNG parameter during beamformer design, the system achieves improved speech intelligibility without excessive self-noise amplification

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The system dynamically adjusts beamforming parameters including WNG constraints and directivity indices based on environmental conditions. The processor modifies array processing parameters in real-time to balance speech enhancement needs against self-noise generation, adapting to different acoustic environments

Inventive Principle:
Principle #15Dynamics

2Measurement precision

If head-mounted arrays are used to provide high directivity, then speech intelligibility improves, but device size and obtrusiveness increase

Engineering Contradiction:
Improvespeech intelligibilityVSAvoidarray size
Core Design Contradiction:
Measurement precisionVSVolume of moving object

Solution Approach 1:

The patent transitions from head-mounted three-dimensional array configurations to planar or linear array arrangements positioned at a distance from the user's head. This dimensional repositioning maintains effective directivity for speech enhancement while dramatically reducing the physical volume and obtrusiveness of the device

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Ease of operation

If off-head arrays are used to reduce device obtrusiveness, then comfort improves, but spatial cues and binaural hearing benefits are lost

Engineering Contradiction:
Improvedevice comfortVSAvoidspatial cues
Core Design Contradiction:
Ease of operationVSLoss of information

Solution Approach 1:

The patent uses digital signal processing as an intermediary to synthesize binaural spatial cues from the planar array microphone signals. The processor applies HRTF-based filtering and spatial rendering techniques to recreate natural binaural hearing effects, including interaural time differences and level differences, thereby compensating for the physical separation between microphones and ears

Inventive Principle:
Principle #24Intermediary (Mediator)

4Ease of manufacture

If free-field beamforming is applied to head-mounted arrays, then simple two-element array performance is achieved, but performance degrades due to head acoustic effects

Engineering Contradiction:
Improvearray design simplicityVSAvoidarray performance
Core Design Contradiction:
Ease of manufactureVSReliability

Solution Approach 1:

The patent applies different processing strategies to different frequency ranges and spatial regions. The system uses frequency-dependent beamforming where low-frequency signals benefit from omnidirectional processing while high-frequency signals use directional beamforming, adapting to the local acoustic characteristics at each frequency band and spatial location

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3105942B1Conversation assistance system
Publication Date: 2018.07.25 BOSE CORP
  • EP3105942B1 patent drawingFigure 1
  • EP3105942B1 patent drawingFigure 2A
  • EP3105942B1 patent drawingFigure 2B

AI summary

A conversation assistance system (16) with a bi-lateral array of microphones (20-23) arranged externally of a space that does not include any array microphones, where the space has a left side, a right side, a front and a back, the array (20-23) comprising a left side sub-array of multiple microphones (20, 21) and a right side sub-array of multiple microphones (22, 23), where each microphone (20-23) has a microphone output signal, and a processor (1 10) that creates from the microphone output signals a left-ear audio signal and a right-ear audio signal. The left-ear audio signal is created based on the microphone output signals from one or more of the microphones of the left-side sub-array (20, 21) and one or more of the microphones of the right-side sub-array (22, 23) and the right-ear audio signal is created based on the microphone output signals from one or more of the microphones of the left-side sub-array (20, 21) and one or more of the microphones of the right-side sub-array (22, 23).