Fractional Delay Filter for Hands-Free Noise Cancellation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing speech processing technologies face challenges in noisy environments, particularly with 'hands-free' devices where background noise interferes with speech signals, making it difficult to extract the user's voice effectively, especially when microphones are close together or in varying noise conditions.

Innovation Solution

The use of two closely spaced microphones with an adaptive filter combiner and a voice activity detector to separate speech signals from noise, employing a fractional delay filter and least mean square (LMS) algorithm for noise reduction, and incorporating a physiological sensor for enhanced voice activity detection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Volume of moving object

If two microphones are placed close together to reduce device size, then device compactness is improved, but noise separation capability deteriorates

Engineering Contradiction:

Improvedevice sizeVSAvoidnoise separation capability

Core Design Contradiction:

Volume of moving objectVSMeasurement precision

Solution Approach 1:

The patent applies fractional delay filtering to adjust the temporal parameters of the signals from closely spaced microphones. By introducing precise sub-sample delays and using interpolation techniques, the system compensates for the reduced spatial separation, maintaining effective noise cancellation capability despite the compact microphone spacing.

Inventive Principle:

Principle #35Parameter changes

2Object-affected harmful factors

If conventional adaptive filtering is used for noise reduction, then noise cancellation is improved, but speech signal distortion occurs

Engineering Contradiction:

Improvenoise cancellationVSAvoidspeech signal integrity

Core Design Contradiction:

Object-affected harmful factorsVSReliability

Solution Approach 1:

The patent performs voice activity detection and filter freezing during speech periods. When speech is detected, the adaptive filter parameters are frozen to prevent distortion of the speech signal, while noise cancellation continues through a different mechanism. This preliminary detection and conditional freezing preserves speech integrity while maintaining noise reduction.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system continuously monitors the speech signal for voice activity and uses this feedback to dynamically adjust the filter operation. When speech is present, the filter is frozen to avoid distortion; when speech is absent, the filter adapts to maximize noise cancellation. This feedback mechanism ensures speech signal integrity is maintained.

Inventive Principle:

Principle #23Feedback

3Measurement precision

If microphone spacing is increased to improve noise rejection, then noise separation capability is improved, but device complexity increases

Engineering Contradiction:

Improvenoise separation capabilityVSAvoidmicrophone arrangement complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts and compensates for the spatial separation effect through signal processing rather than physical separation. By using fractional delay filters to synthetically create the time delays that would naturally occur with larger spacing, the system achieves equivalent noise rejection performance without the physical complexity of widely spaced microphones.

Inventive Principle:

Principle #2Taking out (Extraction)

4Speed

If standard sampling rate is used for processing, then processing speed is improved, but fractional delay accuracy deteriorates

Engineering Contradiction:

Improveprocessing speedVSAvoiddelay precision

Core Design Contradiction:

SpeedVSMeasurement precision

Solution Approach 1:

The patent moves from discrete sample-level delays to continuous fractional delay domains by using interpolation techniques. This dimensional transformation allows the system to achieve sub-sample precision in delay adjustment while maintaining compatibility with standard sampling rates, effectively adding a fractional dimension to the delay control.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

Data Source

PatentUS8682658B2Audio equipment including means for de-noising a speech signal by fractional delay filtering, in particular for a “hands-free” telephony system

Publication Date: 2014.03.25 PARROT FAURECIA AUTOMOTIVE SAS

AI summary

The equipment comprises two microphones, sampling means, and de-noising means. The de-noising means are non-frequency noise reduction means comprising a combiner having an adaptive filter performing an iterative search seeking to cancel the noise picked up by one of the microphones on the basis of a noise reference given by the other microphone sensor. The adaptive filter is a fractional delay filter modeling a delay that is shorter than the sampling period. The equipment also has voice activity detector means delivering a signal representative of the presence or the absence of speech from the user of the equipment. The adaptive filter receives this signal as input so as to enable it to act selectively: i) either to perform an adaptive search for the parameters of the filter in the absence of speech; ii) or else to “freeze” those parameters of the filter in the presence of speech.