Adaptive Directional Hearing Aid Beamformer for Multiple Target Positions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional hearing devices struggle with noise reduction when sound sources are not localized to a single position, as they assume a target sound comes from a specific direction, failing to effectively manage multiple sound sources or uncertain target positions.
Innovation Solution
The implementation of a Generalized EigenVector beamformer (GEV) that maximizes the signal-to-noise ratio (SNR) by determining beamformer weights for a plurality of target positions, using a generalized eigenvalue approach and updating covariance matrices based on voice activity detection and multiple steering vectors, allowing for adaptive noise reduction across various sound sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional beamforming assumes a single target position, then the noise reduction is simplified and computationally efficient, but the system fails to effectively manage multiple sound sources or uncertain target positions
Solution Approach 1:
The patent segments the target positions into a discrete set of K possible positions. Instead of treating the continuous spatial domain, the system divides it into discrete segments (target positions), allowing the beamformer to selectively enhance signals from multiple specific directions while maintaining computational efficiency through discrete position handling.
Solution Approach 2:
The patent implements dynamic adaptability by allowing the beamformer weights to be continuously adjusted based on the estimated target position. The system dynamically switches between different beamforming configurations corresponding to different target positions, enabling it to adapt to changing acoustic environments with multiple sound sources.
2Reliability
If beamformer weights are optimized for a single target position, then the signal-to-noise ratio is maximized for that position, but the system cannot simultaneously optimize for multiple target positions
Solution Approach 1:
The patent merges multiple beamforming operations into a unified framework. By combining the beamforming weight calculations for K different target positions into a single optimization problem, the system can simultaneously optimize signal-to-noise ratio for multiple positions. The combined beamformer weights are derived by integrating the individual position-specific weights through a unified mathematical formulation.
Solution Approach 2:
The patent creates a universal beamforming solution that functions for multiple target positions simultaneously. The developed beamformer structure serves as a multi-functional system that can handle any combination of the K predefined target positions, making it universally applicable to various acoustic scenarios without requiring separate optimization for each position.
3Ease of operation
If the system uses a fixed beamforming configuration, then the implementation is simple and computationally efficient, but it cannot adapt to changing acoustic environments or uncertain target positions
Solution Approach 1:
The patent performs preliminary action by pre-defining K discrete target positions and pre-calculating the beamforming weights for each position. This preparation work is done in advance, allowing the system to quickly adapt to changing environments by simply selecting or combining the pre-computed weights corresponding to the current target positions, rather than performing complex real-time optimization.
Solution Approach 2:
The patent implements feedback mechanisms by continuously estimating the target position from the microphone signals and using this estimation to adjust the beamforming weights. The system monitors the acoustic environment, detects changes in target position, and dynamically updates the beamforming configuration accordingly, creating a closed-loop adaptive system.
Data Source
AI summary
A hearing aid including a multitude M≥2 microphones adapted for providing M electric input signals (x) representative of an environment of a user, at least one beamformer for generating at least one beamformed signal in dependence of beamformer weights (w) configured to be applied to said electric input signals, thereby providing said at least one beamformed signal (Y) as a weighted sum of the M of electric input signals. The beamformer weights (w) are adaptively optimized to a plurality of target positions (θ) by maximizing a target signal to noise ratio (SNR) for sound from the target positions (θ). The signal to noise ratio may be determined in a number of different ways, e.g. in dependence of first and second output variances (|YT2, |YV|2) of said beamformer, when said electric input signals (x) or said beamformed signal (Y) are/is labelled as target (T) and noise (V), respectively.