Hearing Aid Multi-Classifier Acoustic Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern hearing devices face challenges in accurately classifying complex listening situations due to limitations in mapping all noise combinations, leading to potential misclassifications and suboptimal signal processing.
Innovation Solution
A hearing device method utilizing at least three independent classifiers to analyze specific acoustic dimensions, with each classifier receiving a tailored selection of features to generate information about the associated hearing situations, allowing for adaptive adjustment of signal processing algorithms to improve recognition and flexibility in handling multiple, potentially parallel, hearing situations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single classifier is used to analyze listening situations, then the device complexity is reduced, but the measurement precision of acoustic dimension analysis deteriorates
Solution Approach 1:
The patent divides the listening situation analysis into multiple independent classifiers, each responsible for analyzing specific acoustic dimensions (e.g., one classifier for speech detection, another for music detection, another for noise detection). This segmentation allows each classifier to specialize in particular acoustic features, improving measurement precision without requiring a single overly complex classifier system.
2Adaptability or versatility
If multiple classifiers are used to analyze different acoustic dimensions, then the adaptability to various listening situations is improved, but the device complexity increases
Solution Approach 1:
The patent designs multiple classifiers that work together in a unified system, where each classifier serves a specific acoustic dimension but all contribute to the overall listening situation assessment. This multi-functional approach enables the system to adapt to diverse listening situations (speech, music, noise, etc.) while maintaining a structured and manageable classifier architecture rather than requiring completely separate analysis systems for each scenario.
Data Source
Figure 1~2
Figure 3~4
AI summary
In a method according to the invention for operating a hearing device (1) comprising at least one microphone (3) for converting ambient sound into a microphone signal (SM), a number of features (ME, MO, MT, MP, MW, MZ, MM) are derived from the microphone signal (SM) or an input signal (SE) generated therefrom. A specifically assigned selection of these features (ME, MO, MT, MP, MW, MZ, MM) is fed to at least three classifiers (KS, KM, KF), each of which is implemented independently for the analysis of an assigned acoustic dimension. Each classifier (KS, KM, KF) generates information about a specific characteristic of the acoustic dimension assigned to it.Depending on at least one of the at least three pieces of information about the respective characteristic of the associated acoustic dimension, at least one signal processing algorithm (A1, A2, A3, A4), which is used to process the microphone signal (SM) or the input signal (SE) into an output signal (SA), is then modified.