Voice Activation System Power Reduction via Multi-State Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Speech recognition systems in wireless and mobile devices waste significant power as they remain constantly active to monitor ambient environments for speech, leading to substantial power consumption, even when no speech is detected.
Innovation Solution
A voice activation system with multiple stages, where a first stage compares energy characteristics of audio signals to thresholds, activating a second stage only when necessary, which then transitions to a fully-operational state to recognize specific wake-up words, thereby minimizing the active time of power-consuming components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the speech recognition engine remains constantly active to monitor ambient environment for speech, then speech recognition responsiveness is improved, but power consumption increases
Solution Approach 1:
The speech recognition system is divided into multiple operational states: a low-power state where only specific trigger words are detected, and a fully-operational state where the complete vocabulary is recognized. This segmentation allows the system to maintain responsiveness for critical triggers while consuming minimal power during idle periods.
Solution Approach 2:
The speech recognition engine dynamically transitions between operational states based on detected speech patterns. When a trigger word is detected in low-power state, the system transitions to fully-operational state to handle the complete vocabulary, optimizing the balance between responsiveness and power consumption in real-time.
2Use of energy by moving object
If the speech recognition engine operates in low power state detecting only specific trigger words, then power consumption is reduced, but speech recognition capability is limited
Solution Approach 1:
Specific trigger words are pre-designated and configured in the low-power state. These predetermined triggers are prepared in advance to activate the full speech recognition capability when needed, allowing the system to maintain versatility on demand while operating efficiently during idle periods.
Solution Approach 2:
The speech recognition engine is designed to perform multiple functions across different operational states: in low-power state it detects specific triggers, and in fully-operational state it recognizes the complete vocabulary. This multi-functionality allows a single system to adapt its capability level based on power availability and detection needs.
3Use of energy by moving object
If multi-state implementation is used to save power, then power savings are achieved, but device complexity increases
Solution Approach 1:
The system is segmented into distinct operational states with clearly defined transition conditions. Each state has specific components activated only when needed, and transitions between states are triggered by well-defined speech detection events, reducing the complexity of state management compared to continuous full-operation systems.
Data Source
AI summary
A voice activation system is provided. The voice activation system includes a first stage configured to output a first activation signal if at least one energy characteristic of a received audio signal satisfies at least one threshold and a second stage configured to transition from a first state to a second state in response to the first activation signal and, when in the second state, to output a second activation signal if at least a portion of a profile of the audio signal substantially matches at least one predetermined profile.