Hearing System Acoustic Scene Classification Using Directional Channels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern hearing aid devices face challenges in automatically adapting to varying acoustic environments, making it difficult for users to determine the optimal hearing program for speech intelligibility, as existing methods for automatic classification of acoustic scenes are not always effective in providing the best hearing experience.
Innovation Solution
A hearing system that uses an input unit, characterizing unit, and evaluating unit to derive sound-characterizing data and directional information, which are then used to adjust transfer function parameters, allowing for improved automatic adaptation to acoustic environments by linking sound characterization with directional information, enabling more precise determination of acoustic scenes and correct recognition of sound sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If automatic classification of acoustic scenes is implemented using existing methods, then the hearing system can adapt to varying acoustic environments, but the speech intelligibility and accuracy of acoustic scene determination are not sufficient
Solution Approach 1:
The acoustic scene analysis is segmented into multiple independent directional channels, each processing sound from a specific direction. This allows the system to analyze and classify acoustic scenes from different directions separately, improving accuracy without requiring a single complex omnidirectional classifier
Solution Approach 2:
The system adds the directional dimension to acoustic scene classification by obtaining audio signals with different directional characteristics. This transforms the classification problem from a single-dimensional scene type identification to a multi-dimensional analysis incorporating directionality, thereby improving measurement precision
2Adaptability or versatility
If multiple hearing programs are provided for different acoustic environments, then the hearing system can be optimized for various situations, but users have difficulty determining which program is best and switching becomes cumbersome
Solution Approach 1:
The hearing system automatically performs acoustic scene determination and selects appropriate hearing programs without user intervention. The system self-services by processing directional audio signals, determining the acoustic scene, and adjusting parameters automatically, eliminating the burden of manual program selection while maintaining adaptability
Solution Approach 2:
The system uses feedback from directional sound analysis to automatically adjust hearing program parameters. By continuously monitoring the acoustic environment through directional microphones and adjusting settings based on the determined scene type, the system adapts to different environments while keeping operation simple for users
3Measurement precision
If sound sources are recognized without directional information, then the classification process is simpler, but the accuracy of identifying and adapting to specific sound sources and their locations is reduced
Solution Approach 1:
The system applies local quality by analyzing sound characteristics specifically in different directional regions. Each directional channel processes audio signals with its own characteristic directionality, allowing the system to identify sound sources and their locations by comparing patterns across different directional zones, thereby improving recognition accuracy through spatial differentiation
Data Source
Figure 1~2
Figure 3~6
Figure 4~5
AI summary
The invention relates to a method for operating a hearing system (1) comprising an input unit (10), an output unit (80) and a transmission unit (20) operationally interconnecting said input (10) output (80) units. Said transmission unit (20) implements a transfer function (G) describing, how audio signals (S1; R1) generated by said input unit (10) are processed in order to derive audio signals (7) fed to said output unit (80), and can be adjusted by one or more transfer function parameters. Said method comprises obtaining, by means of said input unit (10) and with a first directional characteristic (P1), first audio signals (S1) from incoming acoustic sound (6); deriving from said first audio signals (S1) a first set (C1) of sound-characterizing data; and deriving, in dependence of - first directional information (D1), which is data comprising information on said first directional characteristic (P1), and of - said first set (C1) of sound-characterizing data, a value (T) for each of said one or more transfer function parameters. This allows to gain insight into the acoustic environment and allows for better automatic adjustments of said transfer function (G).