Dynamic Microphone Mixing for Smooth Speaker Switching Noise

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing multi-microphone systems fail to effectively combine signals from different speakers in noisy environments, particularly in scenarios like car telephony, where varying background noise and speech levels lead to noticeable noise jumps and increased noise levels during speaker switching.

Innovation Solution

A dynamic signal mixer system that uses spectral preprocessing to equalize signal-to-noise ratios and noise characteristics across channels, employing a noise reduction algorithm and automatic gain control to adjust signal levels and attenuate background noise based on noise power spectral density and voice activity detection, ensuring smooth transitions between speakers.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If hard switching between active speakers is done, then speaker changes are rapid, but noise jumps and coloration become noticeable

Engineering Contradiction:
Improvespeaker switching speedVSAvoidnoise jumps and coloration
Core Design Contradiction:
SpeedVSObject-affected harmful factors

Solution Approach 1:

The patent implements dynamic mixing ratios that continuously adapt based on voice activity detection and noise level analysis. Instead of fixed hard switching, the system dynamically adjusts the contribution of each microphone signal to the output, allowing smooth transitions that prevent noticeable noise jumps while maintaining rapid speaker changes.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the mixing parameter (mixing ratio) dynamically based on detected voice activity and noise characteristics. By adjusting the weight of each microphone signal according to real-time conditions, the system achieves both rapid speaker switching and smooth noise transitions without audible artifacts.

Inventive Principle:
Principle #35Parameter changes

2Stability of the object's composition

If soft mixing functions are used, then speaker transitions are smooth, but the resulting noise level increases

Engineering Contradiction:
Improvespeaker transition smoothnessVSAvoidresulting noise level
Core Design Contradiction:
Stability of the object's compositionVSObject-affected harmful factors

Solution Approach 1:

The patent applies different processing to different frequency bands and time segments. By analyzing noise characteristics locally in time and frequency domains, the system can apply aggressive noise reduction where needed while maintaining smooth mixing where speech is present, thus achieving both transition smoothness and noise control.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system performs preliminary noise characterization and voice activity detection before mixing occurs. By pre-identifying noise components and speech segments, the system can pre-calculate optimal mixing ratios that ensure smooth transitions without incorporating excessive noise from inactive microphones.

Inventive Principle:
Principle #10Preliminary action

3Manufacturing precision

If spectral preprocessing is applied, then noise characteristics are equalized, but processing complexity increases

Engineering Contradiction:
Improvenoise characteristic equalizationVSAvoidsignal processing complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent divides the signal processing into distinct stages: voice activity detection, noise level estimation, spectral preprocessing, and mixing ratio calculation. By segmenting the complex processing task into modular stages, the system achieves precise noise equalization while making the overall complexity manageable through organized processing steps.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP2647223B1Dynamic microphone signal mixer
Publication Date: 2019.08.07 CERENCE OPERATING CO
  • EP2647223B1 patent drawingFigure 1
  • EP2647223B1 patent drawingFigure 2(a)~3
  • EP2647223B1 patent drawingFigure 4

AI summary

A system and method of signal combining that supports different speakers in a noisy environment is provided. Particularly for deviations in the noise characteristics among the channels, various embodiments ensure a smooth transition of the background noise at speaker changes. A modified noise reduction (NR) may achieve equivalent background noise characteristics for all channels by applying a dynamic, channel specific, and frequency dependent maximum attenuation. The reference characteristics for adjusting the background noise may be specified by the dominant speaker channel. In various embodiments, an automatic gain control (AGC) with a dynamic target level may ensure similar speech signal levels in all channels.