Keystroke Noise Cancellation Using Reference Microphone

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing audio and video conferencing systems face challenges in suppressing keyboard typing noise, which is disruptive and difficult to remove without introducing perceivable distortions, especially due to the spatial proximity of microphones to keyboards and the non-stationary nature of keystroke transients.

Innovation Solution

A semi-supervised acoustic keystroke transient cancellation system using a broadband adaptive MIMO filtering approach with a synchronous reference microphone embedded in the keyboard, which captures keystroke noise unaffected by voice signals, and employs adaptive FIR filters to suppress keystroke noise effectively.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If adaptive noise cancellation is used to reduce keyboard typing noise, then keystroke noise is suppressed, but speech signals may be distorted or attenuated

Engineering Contradiction:
Improvekeystroke noiseVSAvoidspeech signal integrity
Core Design Contradiction:
Object-affected harmful factorsVSReliability

Solution Approach 1:

A reference microphone is introduced as an intermediary element positioned near the keyboard to capture keystroke transient signals separately. This reference signal serves as a mediator that provides clean keystroke information to the adaptive filter, enabling noise cancellation without directly processing and potentially distorting the main speech signal path

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements feedback through adaptive filtering where the reference microphone signal is continuously processed and fed back to cancel keystroke noise in the main microphone signal. The adaptive nature allows the system to learn and adjust to changing acoustic conditions while maintaining speech signal integrity through controlled feedback mechanisms

Inventive Principle:
Principle #23Feedback

2Volume of moving object

If microphones are positioned close to the keyboard for compact device design, then device size is reduced, but keystroke noise capture is increased

Engineering Contradiction:
Improvedevice sizeVSAvoidkeystroke noise
Core Design Contradiction:
Volume of moving objectVSObject-generated harmful factors

Solution Approach 1:

The reference microphone acts as an intermediary that specifically captures keystroke noise generated by the compact keyboard arrangement. By positioning this dedicated reference microphone near the keyboard, the system isolates the harmful keystroke signals without requiring the main speech microphones to be positioned far away, thus maintaining compact device design while managing noise

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The audio capture function is segmented into two separate pathways: one for speech capture and another for keystroke noise capture via the reference microphone. This segmentation allows the system to handle speech and noise independently, enabling compact positioning while using signal processing to separate and eliminate keystroke noise from the speech signal

Inventive Principle:
Principle #1Segmentation

3Object-affected harmful factors

If aggressive noise reduction is applied to remove keystroke transients, then keystroke noise is reduced, but perceivable distortions are introduced in the audio signal

Engineering Contradiction:
Improvekeystroke transient noiseVSAvoidaudio signal fidelity
Core Design Contradiction:
Object-affected harmful factorsVSManufacturing precision

Solution Approach 1:

The adaptive filter using the reference microphone signal as an intermediary provides a controlled method to reduce keystroke transients. By subtracting only the keystroke components identified through the reference signal, the system achieves noise reduction without applying aggressive filtering that would distort the speech signal

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system dynamically changes the parameters of the adaptive filter based on the reference microphone signal characteristics. This allows precise control over the noise cancellation process, adjusting filter parameters to match the keystroke transient characteristics while preserving speech signal fidelity and avoiding perceivable distortions

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3329488B1Keystroke noise canceling
Publication Date: 2019.09.11 GOOGLE LLC
  • EP3329488B1 patent drawingFigure 1
  • EP3329488B1 patent drawingFigure 2(a)~2(d)
  • EP3329488B1 patent drawingFigure 3

AI summary

Provided are methods and systems for acoustic keystroke transient cancellation/suppression for user communication devices using a semi-blind adaptive filter model. The methods and systems are designed to overcome existing problems in transient noise suppression by taking into account some less-defective signal as side information on the transients and also accounting for acoustic signal propagation, including the reverberation effects, using dynamic models. The methods and systems take advantage of a synchronous reference microphone embedded in the keyboard of the user device, and utilize an adaptive filtering approach exploiting the knowledge of this keybed microphone signal.