Hearing Device Own-Voice Detection via Multi-Criteria Voice Detector
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current hearing devices face challenges in accurately detecting own-voice, which is crucial for effective classification of acoustic scenes and efficient processing, especially in multi-purpose usage scenarios like hearing aids and music streaming, leading to potential power drain and interference.
Innovation Solution
A hearing device equipped with a voice detector module that uses multiple microphones and a processor to determine own-voice presence based on various criteria, including power and spectral parameters, and communicates with a contralateral hearing device for improved detection, thereby reducing false positives and negatives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If own-voice detection is implemented using single microphone or simple criteria, then device complexity is reduced, but detection accuracy and reliability deteriorate
Solution Approach 1:
The detection system is segmented into multiple independent voice criteria (first voice criterion, second voice criterion, third voice criterion) that evaluate different aspects of the audio signal. Each criterion can be independently processed and combined, allowing the system to achieve high detection accuracy through multiple specialized checks rather than one complex monolithic detector
Solution Approach 2:
The system transitions from single-microphone detection to multi-microphone spatial detection, adding a spatial dimension to the detection process. By analyzing signals from multiple microphones and evaluating spatial relationships, the system achieves more accurate own-voice detection without proportionally increasing processing complexity
2Reliability
If multiple voice criteria are combined for own-voice detection, then detection reliability is improved, but processing time and computational load increase
Solution Approach 1:
The system performs preliminary analysis of the audio signal to identify potential own-voice conditions before applying full multi-criteria evaluation. By pre-processing signals and identifying promising candidates for detection, the system reduces the computational burden of applying multiple voice criteria, thereby reducing processing time while maintaining high reliability
Solution Approach 2:
The system applies voice criteria selectively rather than always applying all criteria at full strength. In certain conditions, not all three voice criteria need to be fully evaluated, allowing the system to maintain high detection reliability while reducing average processing time through partial evaluation strategies
3Measurement precision
If own-voice detection is made highly accurate, then occlusion effect cancellation and music streaming performance are improved, but power consumption increases
Solution Approach 1:
Instead of continuously applying all voice criteria at full intensity, the system periodically updates detection parameters and adjusts the level of analysis based on current conditions. This periodic action allows the system to maintain high detection precision when needed while reducing power consumption during stable conditions where full analysis is not required
Solution Approach 2:
The system dynamically changes detection parameters such as threshold values, criterion weights, and analysis depth based on operating conditions. By adjusting these parameters, the system can achieve high detection precision in challenging scenarios while consuming less power in favorable conditions, effectively decoupling precision from constant high power consumption
Data Source
AI summary
A hearing device includes: a first microphone and a second microphone for provision of a first microphone input signal and a second microphone input signal, respectively; a voice detector module configured to process the first microphone input signal and the second microphone input signal, the voice detector module configured to detect own-voice of a user of the hearing device; a processor configured to process the first microphone input signal and the second microphone input signal for provision of an electrical output signal based on the first microphone input signal and the second microphone input signal; and a receiver configured to convert the electrical output signal to an audio output signal; wherein the voice detector module is configured to notify a detection of the own-voice to the processor if at least two of a first voice criterion, a second voice criterion, and a third voice criterion are satisfied.