Acoustic Feedback Path Modeling for Hearing Aids
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current hearing assistance devices face challenges with acoustic feedback, particularly due to dynamic acoustic environments, which lead to distortion and whistling sounds, as traditional feedback cancellation methods require a large number of adaptive parameters, resulting in slow convergence and high computational complexity.
Innovation Solution
A method and system that model acoustic feedback paths using a combination of an invariant finite impulse response (FIR) filter and an adaptive FIR filter, with a Bayesian framework for blind deconvolution, incorporating sparsity and exponential decay constraints to reduce the number of adaptive parameters and improve convergence speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional feedback cancellation methods use a large number of adaptive parameters to model acoustic feedback paths, then modeling accuracy is improved, but convergence speed decreases and computational complexity increases
Solution Approach 1:
The patent segments the acoustic feedback path model into two distinct components: a time-invariant common pole-zero filter and time-varying all-zero filters. This segmentation allows the system to use fewer adaptive parameters (only for the all-zero filters) while maintaining modeling accuracy, as the pole-zero filter captures the dominant feedback characteristics without requiring adaptation. The segmentation resolves the contradiction by reducing the number of parameters needing adaptation, thereby improving convergence speed while preserving modeling accuracy.
Solution Approach 2:
The patent changes the parameter representation from traditional all-pole or all-zero models to a hybrid pole-zero model where only specific parameters (the all-zero filter coefficients) are adapted. This parameter change reduces the total number of adaptive parameters from what would be required in traditional approaches, directly improving convergence speed while maintaining or enhancing modeling accuracy through the structured pole-zero representation.
2Measurement precision
If traditional feedback cancellation methods use a large number of adaptive parameters to model acoustic feedback paths, then modeling accuracy is improved, but computational complexity increases
Solution Approach 1:
By segmenting the feedback path model into a fixed pole-zero filter and adaptive all-zero filters, the patent reduces computational complexity. The pole-zero filter requires no adaptation computations, while the all-zero filters use fewer adaptive parameters than traditional approaches. This segmentation maintains modeling accuracy while significantly reducing the computational burden of adaptive filtering operations.
Solution Approach 2:
The patent changes from using many adaptive parameters in traditional models to using fewer adaptive parameters in the all-zero filter component of the pole-zero model. This parameter reduction directly decreases computational complexity while the structured pole-zero form maintains or improves modeling accuracy, resolving the contradiction between accuracy and computational complexity.
3Device complexity
If feedback cancellation uses time-invariant models, then computational complexity is reduced, but adaptability to dynamic acoustic environments deteriorates
Solution Approach 1:
The patent implements dynamics by making the all-zero filter component adaptive while keeping the pole-zero filter time-invariant. This hybrid approach allows the system to adapt to changing acoustic environments through the all-zero filter updates while maintaining the computational efficiency of the fixed pole-zero structure. The adaptive all-zero filters capture environmental variations, resolving the contradiction between simplicity and adaptability.
Solution Approach 2:
The segmentation into fixed pole-zero and adaptive all-zero components allows the system to combine the benefits of both time-invariant and adaptive approaches. The pole-zero filter provides computational efficiency and stability, while the adaptive all-zero filters provide environmental adaptability, resolving the contradiction between computational simplicity and adaptability to dynamic conditions.
Data Source
AI summary
A system and method of determining a filter to cancel feedback signals from input signals in a hearing assistance device includes determining feedback signals for a plurality of feedback paths associated with the device, and determining a model of the plurality of feedback paths, with the model having an invariant portion and a time varying portion. A probable structure of the invariant portion is determined to generate a structural constraint to constrain the plurality of feedback paths, and probability distributions to impose the structural constraint on the invariant portion are determined. During an iterative process, the invariant portion is iteratively determined using the determined probability distributions and the feedback path measurements. A measurement noise variance representative of model mismatch is updated, for each iteration, to reduce a probability of a non-desirable determination of an invariant filter, and the invariant filter is determined in response to a criterion for ending the iterative process being satisfied.