Intermittent Audio Sampling for Low Power Voice Trigger
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for initiating hands-free interaction with mobile devices, such as using pre-arranged activation phrases, lead to increased power consumption and reduced battery life due to continuous audio sampling.
Innovation Solution
A low power voice trigger architecture with a periodic detection window that alternates between active and inactive periods, reducing power consumption by sampling audio only during active periods and powering down components during inactive periods, thereby extending battery life.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If continuous audio sampling is used for voice trigger detection, then voice trigger detection accuracy is improved, but power consumption increases
Solution Approach 1:
The patent implements periodic audio sampling with duty cycling, where the audio front end alternates between active sampling periods and inactive sleep periods. During active periods, audio is sampled at normal rates for voice trigger detection. During inactive periods, sampling is reduced or stopped entirely, allowing significant power savings while maintaining detection capability through periodic activation.
Solution Approach 2:
The system dynamically adjusts the sampling rate and activity state of the audio front end based on detected voice activity. When voice activity is detected during low-power mode, the system transitions to high-power continuous sampling mode. When no voice activity is present, it transitions to low-power intermittent sampling mode, optimizing the balance between detection accuracy and power consumption in real-time.
2Reliability
If continuous audio sampling is used for voice trigger detection, then detection reliability is improved, but battery life decreases
Solution Approach 1:
The audio front end operates in periodic detection windows with duty cycling, alternating between active detection periods and inactive sleep periods. This periodic operation maintains the ability to detect voice triggers reliably during active periods while extending battery life through reduced overall power consumption during inactive periods.
Solution Approach 2:
The system uses voice activity detection (VAD) to preliminarily assess whether voice activity is present before engaging full-power continuous sampling. This preliminary detection allows the system to activate continuous sampling only when necessary, preserving battery life while maintaining detection reliability when voice triggers are actually present.
3Duration of action of stationary object
If intermittent sampling is used to reduce power consumption, then battery life is extended, but detection precision may be reduced
Solution Approach 1:
The system dynamically transitions between low-power intermittent sampling mode and high-power continuous sampling mode based on voice activity detection. When VAD detects voice activity during intermittent sampling, the system immediately switches to continuous sampling mode, ensuring high detection precision is maintained during the critical periods when voice triggers are present while preserving battery life during non-critical periods.
Data Source
AI summary
Systems and methods may provide for using an audio front end of a mobile device to sampled audio from an audio signal during a first portion of a periodic detection window, and reducing a power consumption of one or more components of the audio front end during a second portion of the periodic detection window. Additionally, a determination may be made as to whether voice activity is present in the audio signal based at least in part on the sampled audio. In one example, the length of the first portion and the length of the second portion are defined by a duty cycle of the periodic detection window.