Hearing Aid Ear Canal Microphone Voice Recognition Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current hearing aid technologies face challenges in efficiently and conveniently training voice recognition systems, often requiring predefined speech sequences, additional reference information, and sensitive environmental conditions, which can lead to unsuitable or incorrect adaptations.
Innovation Solution
The method employs an ear canal microphone to analyze signals for the presence of the user's own voice, using adaptive filters and algorithms that adjust based on the ear canal signal, allowing for automatic online training during normal operation without predefined sequences, and adjusts to the acoustic environment, ensuring reliable and effective voice recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional microphones detecting airborne sound are used, then external audio sources can be detected, but the wearer's own voice is perceived differently due to structure-borne sound and occlusion effect
Solution Approach 1:
The patent segments the sound detection function into two separate microphones: one for detecting airborne sound (external sources) and another for detecting structure-borne sound (wearer's own voice). This segmentation allows each microphone to be optimized for its specific detection task, resolving the contradiction between accurate external sound detection and accurate own-voice detection.
Solution Approach 2:
The patent introduces an intermediary signal processing mechanism that combines outputs from two different microphones. The first microphone detects airborne sound while the second detects structure-borne sound, and their signals are processed together to create a composite representation that accurately reflects both external audio sources and the wearer's own voice.
2Measurement precision
If adaptive filters with training procedures are used for voice recognition, then voice recognition accuracy improves, but the system requires predefined speech sequences and additional reference information
Solution Approach 1:
The patent enables the hearing aid system to perform self-service training by automatically adapting filter parameters using real-time signals from the two microphones. The system uses the structural differences between airborne and structure-borne sound to automatically identify and adapt to the wearer's voice characteristics without requiring predefined speech sequences or external reference information.
Solution Approach 2:
The patent dynamically changes filter parameters based on real-time analysis of signals from both microphones. The adaptive filter continuously adjusts its characteristics to optimize voice recognition, changing parameters such as filter coefficients and thresholds based on the detected acoustic environment and voice presence, thereby eliminating the need for static training procedures.
3Adaptability or versatility
If adaptive algorithms are used for online training, then voice recognition adapts to varying environments, but the system may perform incorrect adaptations in unsuitable environmental conditions
Solution Approach 1:
The patent implements a feedback mechanism where the system continuously monitors the acoustic environment using both microphones and adjusts its adaptation behavior accordingly. The feedback loop analyzes signal characteristics to determine whether current environmental conditions are suitable for training, and only performs adaptations when conditions are favorable, thereby preventing incorrect adaptations while maintaining environmental adaptability.
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 approach enables reliable and convenient voice recognition training, reducing the risk of incorrect adaptations and eliminating the need for additional reference information or predefined speech sequences, ensuring effective voice recognition in varying environments.
Implementation Method 1
sound in the wearer's ear canal is detected by the ear canal microphone and a signal corresponding to the sound is output by the ear canal microphone
Implementation Method 2
The audio signal from the first microphone is filtered with a first filter, which attenuates the hearing aid wearer's own voice. Furthermore, the audio signal from the second microphone is attenuated with a second filter, which attenuates an external audio source
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a method for operating a hearing aid (2) which includes an ear canal microphone (6) for detecting sound in the ear canal (8) of the wearer and a voice recognition unit (14) for performing adaptive voice recognition. In this method, sound in the ear canal (8) is detected by the ear canal microphone (6), and a signal (SG) corresponding to the sound is output by the ear canal microphone (6). The adaptive voice recognition training is then controlled solely based on the signal (SG) output by the ear canal microphone (6). The invention further relates to such a hearing aid (2).