Wearable Audio Feedback Detection Via Frequency-Band Phase Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wearable audio devices struggle to efficiently detect and suppress audio feedback, which occurs as uncomfortable whistling sounds due to positive feedback loops between output and input elements, without incurring high computational costs.
Innovation Solution
A method involving time-domain analysis of audio signals to detect phase correlations across frequency bands, generating a feedback detection signal when predefined criteria are met, and selectively suppressing the audio signal at identified feedback frequencies to break the loop.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional audio feedback detection methods are used, then feedback can be detected, but computational cost becomes excessively high
Solution Approach 1:
The audio signal is divided into multiple frequency bands using filter banks, allowing the system to analyze phase correlations in segmented frequency regions rather than processing the entire spectrum uniformly. This segmentation enables efficient detection by focusing computational resources on specific frequency ranges where feedback is most likely to occur.
Solution Approach 2:
The method transforms the audio signal from time domain to frequency domain representation and analyzes phase parameters across different frequency bands. By changing the analysis parameters to focus on phase correlations rather than amplitude alone, the system achieves accurate feedback detection with reduced computational complexity compared to traditional full-spectrum analysis methods.
2Speed
If rapid feedback detection is implemented, then feedback can be suppressed quickly, but computational complexity increases
Solution Approach 1:
The system performs phase correlation analysis on a selected subset of frequency bands rather than analyzing all frequency components. This partial action approach enables rapid detection by focusing only on the most critical frequency regions where feedback is likely to manifest, achieving fast response without the computational burden of complete spectral analysis.
Solution Approach 2:
The audio signal is preprocessed through filter banks that organize frequency components before phase correlation analysis. This preliminary organization of spectral data structures the information in a way that accelerates subsequent feedback detection, allowing the system to quickly identify feedback conditions without performing complex computations on raw audio data.
Data Source
AI summary
The disclosure relates to a method and wearable audio device for detecting an audio feedback condition in an audio signal. The method comprises receiving a series of time frames of a time-domain audio signal, each pair of consecutive time frames of the series of time frames being, at least partly, shifted in time by a time interval; determining in a frequency domain, a phase spectrum for each time frame over a plurality of frequency bands; for each frequency band of the plurality of frequency bands, determining a phase correlation across the determined phase spectra, and generating a pure tone detection signal in case a predefined phase correlation criterion is met.