Hearing Aid Sound Classification via Multi-Device Consensus
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sound environment classification techniques for hearing assistance devices have less than 100% accuracy, leading to potential misclassification and suboptimal parameter settings for users.
Innovation Solution
The use of redundant estimates from multiple hearing assistance devices, including on-the-body and off-the-body devices, to determine a consensus classification through error matrices and error distributions, ensuring accurate environmental classification by comparing and resolving discrepancies between classifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If single device classification is used, then device complexity is low, but classification accuracy is insufficient
Solution Approach 1:
The patent combines classification results from multiple hearing assistance devices (binaural devices, on-the-body devices, off-the-body devices) into a unified classification system. Each device independently classifies the sound environment, and the results are merged through consensus determination to achieve higher accuracy than any single device could achieve alone.
Solution Approach 2:
The system implements feedback mechanisms where classification results and uncertainty values are exchanged between devices via wireless communication. Each device receives classification data from other devices, compares it with its own classification, and uses this feedback to refine the final operational classification through consensus determination.
2Reliability
If multiple devices are used for classification, then classification accuracy improves, but device complexity increases
Solution Approach 1:
The classification system is segmented into independent functional modules: individual device classification modules that operate autonomously, a communication module for data exchange, and a consensus determination module that integrates results. This segmentation allows each component to remain relatively simple while the overall system achieves high reliability through coordinated operation of multiple devices.
3Measurement precision
If consensus determination is implemented, then classification accuracy improves, but processing time increases
Solution Approach 1:
Each device performs preliminary classification independently and simultaneously before consensus determination. The uncertainty values are pre-calculated for each classification result. This preliminary action allows the consensus determination process to focus only on comparing pre-processed results rather than performing full classification analysis, reducing the time penalty of multi-device coordination.
Data Source
AI summary
Techniques are disclosed for classifying a sound environment for hearing assistance devices using redundant estimates of an acoustical environment from two hearing assistance devices and accessory devices. In one example, a method for operating a hearing assistance device includes sensing an environmental sound, determining a first classification of the environmental sound, receiving at least one second classification of the environmental sound, comparing the determined first classification and the at least one received second classification, and selecting an operational classification for the hearing assistance device based upon the comparison.