Wake-Word Microphone Hardware for Low-Power Voice Activation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Voice-controlled devices face significant power consumption issues due to continuous microphone and processor activity when listening for activation phrases, leading to reduced battery life and increased latency, especially in battery-powered devices.
Innovation Solution
A microphone system with dedicated hardware logic for detecting a pre-defined activation phrase, allowing the processor to be switched off and eliminating the need for data connections, enabling local and efficient detection that reduces power consumption and extends battery life.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the processor continuously analyzes audio streams to detect activation phrases, then the detection accuracy is improved, but the power consumption increases significantly
Solution Approach 1:
The audio processing system is segmented into two distinct parts: a low-power dedicated hardware logic circuit that continuously monitors for wake words, and a main processor that remains in sleep mode and only activates when triggered. This segmentation allows continuous monitoring without keeping the power-intensive main processor running, thereby maintaining detection accuracy while significantly reducing power consumption.
Solution Approach 2:
A dedicated hardware logic circuit acts as an intermediary between the microphone and the main processor. This intermediary continuously analyzes audio input and only communicates with the main processor when an activation phrase is detected, eliminating the need for the main processor to continuously process audio streams while ensuring no activation phrase is missed.
2Use of energy by moving object
If the processor is switched off to save power, then power consumption is reduced, but the detection latency increases
Solution Approach 1:
The dedicated hardware logic circuit performs preliminary action by continuously monitoring the audio stream for activation phrases even when the main processor is powered down. This preliminary detection ensures that when the processor needs to be activated, it can do so immediately upon detecting the wake word, minimizing latency while maintaining low power consumption during idle periods.
3Measurement precision
If data connections are used for remote phrase detection, then detection capability is improved, but power consumption and latency increase
Solution Approach 1:
The phrase detection capability is extracted from the main processor and implemented in dedicated hardware logic circuitry within the microphone system. This extraction enables local wake word detection without requiring connection to remote servers, eliminating the power consumption and latency associated with wireless data transmissions while maintaining robust detection capability.
Data Source
AI summary
Methods of low power detection of an activation phrase are described. A microphone system comprises dedicated hardware logic for detecting a pre-defined activation phrase in an audio stream received via a microphone. If the pre-defined activation phrase is detected, the hardware logic sends a trigger signal to activate a module, such as a main speech detection module, which is external to the microphone system and which may be in a low power standby state.