Hearing Aid Speech Separation for Lombard Effect Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hearing aid users experience discomfort and difficulty in conversations due to the Lombard effect, where they inadvertently speak more quietly or loudly in response to amplified self-speech, affecting both their comfort and the clarity of communication.
Innovation Solution
A method for operating a hearing aid system that separates ambient sound into components corresponding to user speech and non-user speech, applying distinct amplification factors to each, adjusting the amplification based on the ratio of these components to optimize speech perception and volume adaptation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the hearing aid amplifies all ambient sound including user speech, then the user's hearing loss is compensated, but the user perceives their own speech as louder than actual and speaks more quietly
Solution Approach 1:
The patent segments the amplified sound into two distinct components: user speech and ambient sound. By separating these components, the system applies different processing to each - specifically, it reduces or eliminates amplification of the user's own speech while maintaining amplification of ambient sounds. This segmentation resolves the contradiction by allowing the user to hear ambient sounds clearly (compensating for hearing loss) without hearing their own speech overly amplified, thus maintaining natural speech volume.
2Reliability
If the hearing aid amplifies user speech, then the user can hear themselves better, but interlocutors have difficulty following the user since the user speaks more quietly
Solution Approach 1:
The system segments sound sources to distinguish between user speech and interlocutor speech. By identifying and selectively processing user speech components, the hearing aid can provide appropriate feedback or amplification for self-monitoring while ensuring that the user maintains natural speech volume for effective communication with others.
Solution Approach 2:
The patent employs feedback mechanisms where the hearing aid monitors the user's own speech output and uses this information to adjust amplification in real-time. This feedback loop allows the system to compensate for the user's reduced speech volume caused by the Lombard effect, thereby improving communication clarity without compromising the user's ability to hear themselves.
3Device complexity
If the hearing aid applies uniform amplification to all sounds, then the implementation is simple, but the user experiences the Lombard effect and speech volume adaptation problems
Solution Approach 1:
The patent implements segmentation of sound sources to distinguish user speech from ambient sounds. This requires more complex processing than uniform amplification, but it enables the system to selectively adjust amplification for different sound types, thereby resolving the speech volume control issues associated with the Lombard effect.
Solution Approach 2:
The system dynamically adjusts amplification factors based on the identified sound source. Rather than applying a static uniform amplification, the hearing aid continuously adapts the amplification level according to whether the detected sound is user speech or ambient sound, resolving the contradiction between processing simplicity and speech volume control.
Data Source
AI summary
A method for operating a hearing aid system includes creating an input signal based on ambient sound. The input signal is divided into a first signal component and a second signal component. The first signal component corresponds to speech of a user and the second signal component does not correspond to speech of the user. A first processed signal is created based on the first signal component and a first amplification factor, and a second processed signal is created based on the second signal component and a second amplification factor. The two processed signals are combined to form an output signal. The first amplification factor is selected depending on a ratio of the first signal component to the second signal component. A hearing aid system and a method for putting a hearing aid system into operation are also provided.