Hearing Device Neural Network ANC Comb-Filter Mitigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hearing aids often suffer from the comb-filter effect due to signal processing delays, which degrade sound quality by introducing echoes or reverberations, and existing ANC technologies are limited in reducing these effects, especially in the frequency domain.
Innovation Solution
A hearing device that integrates time-domain and frequency-domain audio signal processing using an end-to-end neural network for active noise cancellation and advanced audio signal processing, generating a compensation mask to mitigate the comb-filter effect and improve audio quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If digital signal processing is performed in hearing aids to amplify sound, then sound amplification is achieved, but signal processing delay occurs which causes comb-filter effect and degrades sound quality
Solution Approach 1:
The patent divides the audio signal processing into two separate domains: time-domain processing for active noise cancellation (ANC) and frequency-domain processing for advanced audio signal processing. This segmentation allows each domain to be optimized independently, with time-domain processing handling latency-critical ANC tasks and frequency-domain processing handling computationally intensive audio enhancement, thereby reducing overall processing delay while maintaining sound amplification quality
Solution Approach 2:
The patent merges time-domain and frequency-domain processing within a unified hearing device architecture, integrating both processing paths to work together on the same audio input. This combination enables the device to simultaneously perform low-latency ANC operations and advanced frequency-domain audio processing, achieving both speed and quality improvements that neither domain could achieve alone
2Loss of time
If Active Noise Cancellation is performed in time domain, then low latency is achieved, but advanced audio signal processing capabilities are limited
Solution Approach 1:
The patent transitions from single-domain (time-domain only) processing to multi-domain processing by adding the frequency dimension. The system performs ANC in the time domain for low latency, then transforms the signal to the frequency domain using STFT to enable advanced audio signal processing operations. This dimensional expansion allows the system to access sophisticated processing algorithms while maintaining low-latency ANC performance
3Adaptability or versatility
If Active Noise Cancellation is performed in frequency domain using STFT, then advanced audio signal processing is enabled, but signal processing delay increases significantly
Solution Approach 1:
The patent segments the noise cancellation function from the advanced audio processing function. Time-domain ANC operates independently with minimal processing delay to handle latency-critical noise cancellation, while frequency-domain processing handles advanced audio enhancement separately. This segmentation prevents the computationally intensive STFT operations from adding delay to the critical ANC path, maintaining low latency while still enabling advanced processing
Data Source
AI summary
A hearing device is disclosed, comprising a main microphone, M auxiliary microphones, a transform circuit, a processor, a memory and a post-processing circuit. The transform circuit transforms first sample values in current frames of a main audio signal and M auxiliary audio signals from the microphones into a main and M auxiliary spectral representations. The memory includes instructions to be executed by the processor to perform operations comprising: performing ANC over the first sample values using an end-to-end neural network to generate second sample values; and, performing audio signal processing over the main and the M auxiliary spectral representations using the end-to-end neural network to generate a compensation mask. The post-processing circuit modifies the main spectral representation with the compensation mask to generate a compensated spectral representation, and generates an output audio signal according to the second sample values and the compensated spectral representation.