Binaural Hearing System Speech Intelligibility Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current hearing aids are ineffective in improving speech intelligibility in complex acoustic environments with noise sources and reverberation, as they indirectly maximize signal-to-noise ratio rather than directly addressing speech intelligibility.
Innovation Solution
A binaural hearing system that estimates speech intelligibility online by processing acoustic signals from both ears and adjusts signal processing to maximize this estimate, using wireless communication between hearing aids to share and process signals, and incorporates a hearing loss model to provide modified signals for improved intelligibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If hearing aids process microphone signals to maximize signal-to-noise ratio, then speech intelligibility is indirectly improved, but the approach is indirect and implicit without a clear one-to-one map
Solution Approach 1:
The patent implements a feedback mechanism where the speech intelligibility predictor continuously monitors the processed signals from both hearing aids and provides real-time feedback to the signal processing units. This feedback loop allows the system to adjust processing parameters dynamically based on actual intelligibility performance, resolving the contradiction by enabling direct optimization without excessive complexity through adaptive control.
Solution Approach 2:
The system changes processing parameters (such as beamformer weights, noise reduction gains, and frequency weighting) based on the intelligibility estimate. By dynamically adjusting these parameters in response to the predicted intelligibility score, the system achieves direct optimization of speech intelligibility while maintaining manageable processing complexity through parameter-driven adaptation rather than fundamentally complex reconfiguration.
2Ease of manufacture
If hearing aids use indirect noise reduction systems to maximize signal-to-noise ratio, then processing becomes practically possible, but speech intelligibility improvement is insufficient in complex acoustic situations
Solution Approach 1:
The feedback mechanism allows the system to learn from actual intelligibility performance and adjust noise reduction strategies accordingly. This enables practical implementation of intelligent noise reduction that adapts to complex acoustic situations, resolving the contradiction by making the system both practically implementable and effective in challenging environments through continuous adaptation.
Solution Approach 2:
The system dynamically adjusts noise reduction parameters and processing strategies based on real-time intelligibility predictions and acoustic scene analysis. This dynamic adaptation allows the system to maintain practical implementability while significantly improving speech intelligibility in complex acoustic situations through flexible, context-dependent processing.
3Reliability
If the system processes acoustic signals from both ears to produce binaural SI estimate, then speech intelligibility is directly enhanced, but wireless communication requirements increase system complexity
Solution Approach 1:
The hearing aids utilize their existing wireless communication capabilities for dual purposes: both for standard audio transmission and for sharing processed signals between ears to enable binaural processing. This multi-functionality approach allows the system to achieve binaural intelligibility estimation without adding separate dedicated communication hardware, thereby limiting the increase in system complexity.
Solution Approach 2:
Each hearing aid leverages its own local processing resources to compute the binaural intelligibility estimate using signals from both ears, with one hearing aid performing the computation while the other provides input signals. This self-service approach distributes the computational burden and avoids requiring a centralized complex processing system, thus managing overall system complexity while achieving the binaural intelligibility enhancement.
Data Source
AI summary
The application relates to a binaural hearing system comprising left and right hearing devices, e.g. hearing aids, each comprising a) a multitude of input units, each providing a time-variant electric input signal x.sub.i(t) representing sound received at an i.sup.th input unit, t representing time, the electric input signal x.sub.i(t) comprising a target signal component s.sub.i(t) and a noise signal component v.sub.i(t), the target signal component originating from a target signal source; b) a configurable signal processing unit for processing the electric input signals and providing a processed signal y(t); c) an output unit for creating output stimuli to the user, d) transceiver circuitry allowing information to be exchanged between the hearing devices, and e) a binaural speech intelligibility (SI) prediction unit for providing a binaural SI-measure of the predicted speech intelligibility of the user when exposed to said output stimuli, based on processed signals y.sub.l(t), y.sub.r(t) from the signal processing units of the respective left and right hearing devices. This allows the hearing devices to control the processing of the respective electric input signals based on said binaural SI-measure.