Breath Wake-Up Control for Uninterrupted Voice Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for waking up voice assistants, such as button and keyword wake-up technologies, are cumbersome and result in low user experience.
Innovation Solution
A breath wake-up technology is introduced, where an electronic device uses an inertial detection sensor and sound acquisition sensor to detect breath and voice data, and an audio digital signal processor to determine if the data matches preset thresholds, triggering the wake-up of an application through a breath wake-up software module.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If button wake-up technology is used, then the voice assistant can be woken up, but the operation is cumbersome and user experience is low
Solution Approach 1:
The patent replaces the mechanical button press operation with breath-based acoustic detection. The breath wake-up processing apparatus detects breath characteristics (flow rate, pressure changes) through acoustic sensors, eliminating the need for physical button interaction and enabling hands-free wake-up operation.
Solution Approach 2:
The system uses the user's own breath as the trigger signal, making the wake-up process intuitive and automatic. The breath detection mechanism continuously monitors respiratory patterns and automatically triggers wake-up when specific breath characteristics are detected, without requiring additional user actions.
2Ease of operation
If keyword wake-up technology is used, then the voice assistant can be woken up, but the process is cumbersome and user experience is low
Solution Approach 1:
The patent replaces keyword-based acoustic recognition with breath-specific acoustic detection. Instead of recognizing spoken words, the system detects unique acoustic characteristics of breath (flow patterns, pressure variations), providing a more direct and faster wake-up mechanism.
Solution Approach 2:
The system focuses on detecting specific local characteristics of breath (acoustic impedance, flow rate variations, pressure changes) rather than requiring full sentence recognition. This localized detection approach enables faster and more reliable wake-up response.
3Speed
If the breath wake-up processing apparatus continuously detects breath wake-up, then the application can be quickly woken up, but the current voice recognition may be interrupted
Solution Approach 1:
The system dynamically adjusts the breath detection state based on application status. When the application is in sleep state, breath detection is active for quick wake-up. When the application is running and needs voice recognition, the system temporarily suspends breath detection to prevent interruptions, then resumes detection when appropriate.
Solution Approach 2:
The breath wake-up software module receives feedback from the first application about its operational state and adjusts breath detection accordingly. The system monitors application status signals and modulates breath detection activity to maintain voice recognition continuity while preserving quick wake-up capability.
4Reliability
If the breath wake-up processing apparatus stops detecting breath wake-up when starting the application, then the current voice recognition is not interrupted, but the next wake-up function is not prepared
Solution Approach 1:
The system performs preliminary breath detection during the application startup phase before voice recognition begins. This preliminary detection prepares the breath wake-up function for the next wake-up event while the application is initializing, ensuring both current voice recognition quality and future wake-up readiness.
Solution Approach 2:
The breath detection function maintains continuous operational readiness through staged activation. Detection is activated in phases: fully active during sleep state for quick wake-up, partially active during startup preparation, and suspended during active voice recognition. This phased continuity ensures both current operation quality and future wake-up readiness.
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 method ensures timely and convenient application wake-up, prevents interruption of voice recognition, and prepares for the next wake-up, thereby improving user experience.
Implementation Method 1
uses an inertial detection sensor and sound acquisition sensor to detect breath and voice data
Implementation Method 2
uses an audio digital signal processor to determine if the data matches preset thresholds
Data Source
AI summary
The method includes: A breath wake-up processing apparatus sends voice data in first data when detecting that the obtained first data is used for indicating to wake up a first application through breath. A breath wake-up software module stores the voice data, starts the first application, and controls the breath wake-up processing apparatus to stop detecting breath wake-up of the first application. The first application sends a first notification when successfully calling the breath wake-up software module. The breath wake-up software module sends the voice data to the first application. The first application performs voice recognition on the voice data. The first application sends a second notification when determining, based on the voice data, that the voice recognition ends. The breath wake-up software module controls, in response to the second notification, the breath wake-up processing apparatus to start detecting next breath wake-up.