Trigger Phrase Validation via Energy and Voice Activity Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Always-on audio systems face challenges in accurately detecting trigger phrases due to high False Accept rates, especially in noisy environments, which degrades the user experience.
Innovation Solution
An electronic device with a processor that receives audio signals, measures characteristics such as average energy, maximum energy, and voice active frames of detected trigger phrases, and compares these to threshold values based on analysis of valid trigger phrase instances to determine validity, thereby reducing False Accepts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the system continuously listens for trigger phrases in always-on mode, then the user experience is simplified and manual activation is eliminated, but the False Accept rate increases especially in noisy environments
Solution Approach 1:
The system performs preliminary actions by continuously analyzing audio characteristics and maintaining voice activity detection states before a trigger phrase is fully detected. This allows the system to prepare validation metrics in advance, enabling faster and more accurate False Accept detection when a potential trigger is heard, thus resolving the contradiction between continuous listening and reliability.
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring voice activity detection metrics, energy levels, and temporal patterns during the trigger phrase detection process. This real-time feedback allows the system to dynamically adjust its detection sensitivity and validate trigger phrases more accurately, reducing False Accepts while maintaining ease of operation.
2Device complexity
If the system uses simple trigger phrase detection, then the device complexity is reduced, but the measurement precision of trigger phrase validity decreases
Solution Approach 1:
The detection system is segmented into multiple independent analysis components: voice activity detection modules, energy analysis modules, temporal pattern analysis modules, and validation decision modules. Each segment processes specific aspects of the audio signal independently, then combines results for final validation. This segmentation improves measurement precision without significantly increasing overall device complexity.
Solution Approach 2:
The system employs universal audio processing components that perform multiple functions: the same voice activity detection infrastructure is used for both continuous background monitoring and trigger phrase validation. Energy analysis serves dual purposes in detecting both speech presence and trigger phrase characteristics. This multi-functionality approach maintains device complexity while improving measurement precision through validated algorithms.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An electronic device includes a microphone (108) that receives an audio signal, and a processor that is electrically coupled to the microphone (108). The processor (204, 300) detects a trigger phrase in the received audio signal and measure characteristics of the detected trigger phrase. Based on the measured characteristics of the detected trigger phrase, the processor (204,300) determines whether the detected trigger phrase is valid.