Low-Power Accelerometer Voice Activity Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice activity detection (VAD) techniques in audio headsets, such as true wireless stereo (TWS) headphones, consume high power due to the use of microphones and bone conduction accelerometers, making them unsuitable for portable devices with limited power supplies.
Innovation Solution
A device and method utilizing a low-cost, low-power conventional accelerometer to detect voice activity by generating an acceleration signal, filtering it, and classifying the presence or absence of speech using machine learning techniques, thereby reducing the need for a separate microcontroller and minimizing power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a microphone and bone conduction accelerometer are used for voice activity detection, then voice detection accuracy is improved, but power consumption increases significantly
Solution Approach 1:
The patent extracts the voice activity detection function from the traditional microphone-based system and implements it using a conventional accelerometer. By removing the microphone and bone conduction accelerometer while retaining the VAD capability through accelerometer data processing, the solution eliminates the high power consumption components while maintaining detection functionality.
Solution Approach 2:
The patent replaces the acoustic mechanical system (microphone capturing sound waves) with a mechanical vibration-based system (accelerometer capturing bone vibrations). This substitution allows the system to detect voice activity through mechanical vibrations transmitted through the skull and jawbone, eliminating the need for high-power acoustic sensing components.
2Adaptability or versatility
If a bone conduction accelerometer is used for voice activity detection, then voice detection capability is improved, but device complexity and power consumption increase
Solution Approach 1:
The patent makes the conventional accelerometer multi-functional by using it for both motion detection (existing function) and voice activity detection (new function). This eliminates the need for separate dedicated VAD components, reducing device complexity while maintaining versatile detection capabilities.
Solution Approach 2:
The patent removes the bone conduction accelerometer from the system while retaining voice detection capability through standard accelerometer data processing. This extraction of the specialized component simplifies the device architecture while preserving the essential VAD function.
3Measurement precision
If high-rate data processing is performed for voice activity detection, then detection accuracy is improved, but power consumption increases
Solution Approach 1:
The patent applies partial action by processing accelerometer data at optimized intervals rather than continuous high-rate processing. By selectively processing data only when necessary (e.g., when motion thresholds are exceeded or during expected speech periods), the system maintains adequate detection accuracy while significantly reducing power consumption.
Solution Approach 2:
The patent changes the data processing parameters by adjusting sampling rates, threshold values, and processing intervals based on operational context. This dynamic parameter adjustment allows the system to maintain detection accuracy while optimizing power consumption according to actual usage conditions.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly reduces overall current consumption, making it ideal for portable audio devices by using a low-power accelerometer to detect voice activity, thereby extending battery life and improving power efficiency.
Implementation Method 1
a bone conduction accelerometer as bone vibration signals propagating through the human body (e.g., bone and tissue vibrations through the jaw or ear canal)
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The device (10) and the method are directed to detect presence or absence of human speech. The device and method utilize a low-power accelerometer (14). The device and method generate an acceleration signal using the accelerometer, filter the acceleration signal with a band pass filter or a high pass filter, determine at least one calculation of the filtered acceleration signal, detect a presence or absence of a voice based on the at least one calculation, and output a detection signal that indicates the presence or absence of the voice. The device and method are well suited for portable audio devices, such as true wireless stereo headphones, that have a limited power supply.