Hearing System Camera-Based Audio Source Identification in Complex Scenes

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Hearing devices often fail to adequately support users in complex acoustic environments due to incorrect settings or misclassification of acoustic environments, making it challenging for users to adjust settings effectively.

Innovation Solution

A hearing system equipped with a camera and display, utilizing artificial intelligence or neural networks to identify audio sources visually, allowing users to easily select and modify sound inputs through a graphical interface, enabling quick adjustments in real-time.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If automatic acoustic environment classification is used, then hearing device settings are adjusted automatically, but misclassification occurs in complex acoustic environments

Engineering Contradiction:
Improveautomatic classificationVSAvoidclassification accuracy
Core Design Contradiction:
Extent of automationVSReliability

Solution Approach 1:

The patent introduces a camera as an intermediary device that captures visual information about the acoustic environment. This visual data serves as a mediator between the complex acoustic scene and the classification algorithm, providing additional cues that help accurately identify audio sources and their locations without requiring perfect acoustic classification alone.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent adds a visual dimension to the traditional acoustic-only classification approach. By incorporating camera images that show the spatial location and visual appearance of audio sources, the system transitions from one-dimensional acoustic analysis to multi-dimensional sensing, improving classification reliability in complex environments.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Adaptability or versatility

If manual adjustment options are provided, then users can customize settings, but the range of modifiers is overwhelming and difficult to navigate

Engineering Contradiction:
Improveadjustment optionsVSAvoiduser interface complexity
Core Design Contradiction:
Adaptability or versatilityVSEase of operation

Solution Approach 1:

The system automatically presents only the relevant adjustment options based on the detected acoustic environment and identified audio sources. Instead of requiring users to navigate through all possible modifiers, the system serves itself by intelligently filtering and presenting only the necessary controls for the current situation, simplifying the user interface while maintaining full adaptability.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent applies different levels of interface complexity to different situations. In simple acoustic environments, minimal controls are presented, while in complex environments, more advanced options become available. This local adaptation of interface quality ensures ease of operation in simple cases while maintaining versatility when needed.

Inventive Principle:
Principle #3Local quality

3Productivity

If visual identification of audio sources is implemented, then users can quickly select and adjust specific sound inputs, but device complexity increases

Engineering Contradiction:
Improveadjustment speedVSAvoidsystem components
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent makes the camera serve multiple functions: it captures images for visual identification of audio sources, determines their spatial location, and provides contextual information for classification. This multi-functionality allows the same component to support multiple features (visual selection, automatic classification, spatial awareness) without proportionally increasing device complexity.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP4178228B1Method and computer program for operating a hearing system, hearing system, and computer-readable medium
Publication Date: 2025.08.27 SONOVA AG
  • EP4178228B1 patent drawingFigure 1
  • EP4178228B1 patent drawingFigure 2
  • EP4178228B1 patent drawingFigure 3~4

AI summary

A method for operating a hearing system (10) is provided. The hearing system (10) comprises a hearing device (12) configured to be worn at an ear of a user, a user device (14) communicatively coupled to the hearing device (12) and comprising a camera (36) and a display (30). The hearing device (12) comprises at least one sound input module (20) for generating an audio signal indicative of a sound detected in an environment of the hearing device (12), a first processing unit (40) for modifying the audio signal, and at least one sound output module (22) for outputting the modified audio signal. The method comprises: receiving image data from the camera (36), the image data being representative for a scene (80) in front of the camera (36); receiving an audio signal from the at least one sound input module (20), the audio signal being representative for the acoustic environment of the hearing device (12) substantially at a time the image data have been captured, wherein the acoustic environment comprises at least one audio source and wherein the audio signal is at least in part representative for a sound from the audio source; determining at least one visual object (88) as the audio source, within the scene (80) from the image data and the audio signal.