Acoustic Voice Activity Detection Using Virtual Microphones

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing noise suppression systems struggle to accurately distinguish voiced and unvoiced speech from background noise, especially in noisy environments, due to reliance on single microphone data and limitations in noise reduction methods.

Innovation Solution

The use of a two-microphone array to form virtual directional microphones with similar noise responses and dissimilar speech responses, employing adaptive and fixed filter techniques to calculate an energy ratio for voice activity detection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If single microphone data is used for voice activity detection, then device complexity is reduced, but measurement precision deteriorates due to noise and signal content uncertainties

Engineering Contradiction:

Improvemicrophone array configurationVSAvoidvoice activity detection accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent segments the audio signal processing into multiple channels by using multiple microphones (at least two) to capture different spatial perspectives of the acoustic environment. Each microphone provides independent signal data that is processed separately through voice activity detection algorithms, allowing the system to segment noise from speech based on spatial and spectral characteristics.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent transitions from single-microphone monaural processing to multi-microphone spatial processing by adding the spatial dimension. By analyzing signals from multiple microphones positioned at different locations, the system creates a multi-dimensional signal space where noise and speech can be distinguished based on their different spatial distributions and propagation characteristics.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Object-affected harmful factors

If noise suppression methods are applied to speech signals, then signal-to-noise ratio is improved, but speech distortion may occur

Engineering Contradiction:

Improvebackground noise levelVSAvoidspeech signal fidelity

Core Design Contradiction:

Object-affected harmful factorsVSManufacturing precision

Solution Approach 1:

The patent implements feedback mechanisms where the output of the voice activity detection and noise suppression processes is continuously monitored and fed back to adjust the processing parameters. The system uses the detected speech presence information to dynamically control the noise suppression intensity, ensuring that suppression is applied only when noise is present and speech is detected, thereby preventing speech distortion.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent dynamically changes processing parameters based on the detected acoustic conditions. The voice activity detection algorithms adjust parameters such as suppression intensity, filter coefficients, and detection thresholds in real-time based on the estimated speech presence and noise characteristics, optimizing the balance between noise reduction and speech preservation.

Inventive Principle:

Principle #35Parameter changes

3Ease of manufacture

If acoustic-only voice activity detection is used, then cost is reduced compared to non-acoustic sensors, but reliability deteriorates in noisy environments

Engineering Contradiction:

Improvesystem costVSAvoidvoice detection reliability in noise

Core Design Contradiction:

Ease of manufactureVSReliability

Solution Approach 1:

The patent introduces virtual microphones as intermediary processing elements that combine signals from physical microphones in specific ways to create enhanced detection channels. These virtual microphones act as mediators that process the raw acoustic signals to emphasize speech characteristics and suppress noise, improving detection reliability without requiring additional physical sensors.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent creates composite signal processing structures by combining outputs from multiple physical microphones and virtual microphones to form enhanced detection signals. This composite approach integrates information from multiple sources with different noise and speech characteristics, creating a more robust detection mechanism that outperforms individual microphones while remaining acoustic-only.

Inventive Principle:

Principle #40Composite materials

Data Source

PatentUS20250097635A1Acoustic voice activity detection (AVAD) for electronic systems

Publication Date: 2025.03.20 JAWBONE INNOVATIONS LLC

AI summary

Acoustic Voice Activity Detection (AVAD) methods and systems are described. The AVAD methods and systems, including corresponding algorithms or programs, use microphones to generate virtual directional microphones which have very similar noise responses and very dissimilar speech responses. The ratio of the energies of the virtual microphones is then calculated over a given window size and the ratio can then be used with a variety of methods to generate a VAD signal. The virtual microphones can be constructed using either an adaptive or a fixed filter.