Hearing Aid Voice Recognition Spatial Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hearing aid users face challenges in recognizing their own voice due to altered perception caused by airborne noise amplification and occlusion effects, leading to unfamiliar and alienated self-voice perception.
Innovation Solution
A method and device for a hearing aid that utilizes two acousto-electric transducers and spatial separation filters to differentiate and attenuate the wearer's own voice from external audio sources, employing adaptive filter parameters determined through training processes to enhance recognition reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If airborne noise is amplified to compensate for hearing loss, then hearing sensitivity is improved, but the wearer's own voice perception becomes altered and unfamiliar
Solution Approach 1:
The audio signal is segmented into different spatial components using multiple microphones and beamforming filters. The system separates the wearer's voice from external sounds by creating distinct signal paths, allowing independent processing of self-voice and external audio to resolve the perception distortion caused by amplification.
Solution Approach 2:
Beamforming filters act as intermediaries that process the raw microphone signals to extract and identify the wearer's own voice. These filters create a mediated representation of self-voice that can be recognized and processed differently from external sounds, bridging the gap between amplified audio and natural perception.
2Object-affected harmful factors
If the auditory canal is closed by the hearing aid to block external noise, then noise protection is improved, but self-voice perception becomes occluded and alienated
Solution Approach 1:
The system extracts the wearer's own voice signal from the mixed audio input by using spatial filtering and beamforming techniques. By taking out the self-voice component separately, the system can process it differently to compensate for occlusion effects while maintaining noise blocking through the physical seal.
Solution Approach 2:
Instead of trying to prevent occlusion effects through physical design, the system inverts the approach by using signal processing to identify and separate self-voice after occlusion has occurred. The beamforming filters reverse the mixing process to recover natural self-voice characteristics despite the closed auditory canal.
3Measurement precision
If multiple microphones and spatial filters are used to differentiate self-voice, then recognition accuracy is improved, but device complexity increases
Solution Approach 1:
The beamforming filters use dynamic adaptation to adjust their parameters based on the acoustic environment and wearer characteristics. This dynamic behavior allows the system to maintain high recognition accuracy across different situations without requiring a fixed complex structure for every possible scenario.
Solution Approach 2:
The system changes filter parameters such as delay times and weighting factors to optimize self-voice recognition for different listening conditions. By adjusting these parameters dynamically rather than using a fixed complex filter structure, the system achieves high accuracy with manageable complexity.
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
The solution enables quick and reliable recognition of the wearer's own voice by effectively distinguishing it from external sources, improving hearing aid user experience by accounting for individual spatial and physiological characteristics.
Implementation Method 1
The apparatus receives audio signals from at least two acousto-electric transducers
Implementation Method 2
first filter parameters of the first filter are determined, with the first filter parameters being designed to attenuate the hearing aid device wearer's own voice
Data Source
Figure 1~2
Figure 3
AI summary
The invention relates to a method for quickly recognizing the user's own voice for a hearing aid device, and to a corresponding device for carrying out the method. In the method, the hearing aid receives audio signals from at least two acousto-electrical transducers. It has a device with a first filter and a second filter for spatial separation. In the method, first filter parameters of the first filter are determined, which attenuate the user's own voice. Furthermore, second filter parameters of the second filter are determined, with the second filter parameters being used to attenuate an external audio source. The device is then operated with the first filter parameters for the first filter and the second filter parameters for the second filter. The user's own voice is recognized depending on an output signal of the first filter and the second filter.