Microphone Decimation Synchronization for Low-Power Voice Detection

Solution Overview

Problem

Existing voice activity detection (VAD) approaches in mobile devices consume significant power, leading to battery life issues and user dissatisfaction, as the algorithms are often not compatible with standardized interfaces in mobile devices, limiting their continuous usage.

Innovation Solution

Integration of VAD or acoustic activity detection (AAD) modules into microphones with application-specific circuitry, along with a standardized interface, allows for reduced power consumption and multiple operational modes, including a low-power mode that detects acoustic events like voice signals, using an external clock signal to determine a division ratio and decimation factor for efficient sampling.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If voice activity detection algorithms are constantly running to enable continuous speech recognition, then speech recognition availability is improved, but power consumption increases significantly

Engineering Contradiction:

Improvespeech recognition availabilityVSAvoidpower consumption

Core Design Contradiction:

Adaptability or versatilityVSUse of energy by moving object

Solution Approach 1:

The system segments the speech recognition functionality into two parts: a low-power acoustic activity detection module that continuously monitors for speech presence, and a full speech recognition algorithm that only activates when speech is detected. This segmentation allows continuous availability of speech recognition without constantly running the power-intensive algorithms.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Instead of continuously running the speech recognition algorithm, the system uses periodic acoustic activity detection to determine when speech is present. The full recognition algorithm is activated only during periods when speech is detected, creating a periodic activation pattern that reduces overall power consumption while maintaining speech recognition availability.

Inventive Principle:

Principle #19Periodic action

2Adaptability or versatility

If voice activity detection algorithms are constantly running to enable continuous speech recognition, then speech recognition availability is improved, but battery life deteriorates

Engineering Contradiction:

Improvespeech recognition availabilityVSAvoidbattery life

Core Design Contradiction:

Adaptability or versatilityVSDuration of action of moving object

Solution Approach 1:

The system segments the speech recognition functionality into two parts: a low-power acoustic activity detection module that continuously monitors for speech presence, and a full speech recognition algorithm that only activates when speech is detected. This segmentation allows continuous availability of speech recognition without constantly running the power-intensive algorithms.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Instead of continuously running the speech recognition algorithm, the system uses periodic acoustic activity detection to determine when speech is present. The full recognition algorithm is activated only during periods when speech is detected, creating a periodic activation pattern that reduces overall power consumption while maintaining speech recognition availability.

Inventive Principle:

Principle #19Periodic action

3Reliability

If non-standardized microphone interfaces are used to achieve optimal performance, then device performance is improved, but compatibility with mobile devices deteriorates

Engineering Contradiction:

Improvedevice performanceVSAvoidcompatibility

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The microphone device is designed with a universal interface that can interface with standardized mobile device connections. The microphone incorporates both acoustic activity detection and speech recognition capabilities in a single device that adheres to standardized interface protocols, allowing it to function optimally while being compatible with various mobile devices.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent combines the microphone, acoustic activity detection module, and speech recognition algorithm into a single integrated device. This merging eliminates the need for separate components with complex interconnections, allowing the device to achieve optimal performance while using standardized interfaces for broad compatibility.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS9711166B2Decimation synchronization in a microphone

Publication Date: 2017.07.18 KNOWLES ELECTRONICS LLC

AI summary

An external clock signal having a first frequency is received. A division ratio is automatically determined based at least in part upon a second frequency of an internal clock. The second frequency is greater than the first frequency. A decimation factor is automatically determined based at least in part upon the first frequency of the external clock signal, the second frequency of the internal clock signal, and a predetermined desired sampling frequency. The division ratio is applied to the internal clock signal to reduce the first frequency to a reduced third frequency. The decimation factor is applied to the reduced third frequency to provide the predetermined desired sampling frequency. Data is clocked to a buffer using the predetermined desired sampling frequency.