Residual Echo Suppression via Frequency-Selective Gain Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic devices face challenges in effectively suppressing residual echo during communication sessions, particularly when both local and remote speech are present simultaneously, leading to suboptimal audio quality due to conventional echo cancellation methods.
Innovation Solution
The system employs a residual echo suppressor that determines double-talk conditions in individual frequency bands, applying dynamic gain values to selectively attenuate or pass audio data, thereby improving echo suppression and maintaining audio quality during double-talk scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If conventional echo cancellation methods are used, then echo suppression is achieved, but audio quality deteriorates during double-talk scenarios
Solution Approach 1:
The patent divides the audio signal processing into frequency-domain segments (frequency bins), allowing independent processing of different frequency components. This segmentation enables selective suppression of echo in specific frequency ranges while preserving speech content in other ranges, resolving the contradiction between echo suppression and audio quality maintenance during double-talk scenarios.
Solution Approach 2:
The patent applies different processing strategies to different frequency bins based on local characteristics. By analyzing correlation values individually for each frequency bin and applying frequency-selective gain adjustments, the system achieves localized optimization where echo is suppressed in affected frequency ranges while speech quality is preserved in other ranges, thereby resolving the contradiction between echo suppression and overall audio quality.
2Object-affected harmful factors
If aggressive echo suppression is applied, then residual echo is reduced, but speech distortion increases
Solution Approach 1:
The patent employs dynamic gain adjustment based on real-time correlation analysis. The system continuously monitors correlation values between echo reference signals and microphone signals, and adaptively adjusts frequency-selective gain values accordingly. This dynamic approach allows aggressive suppression when echo is dominant while maintaining speech fidelity when speech content is present, resolving the contradiction between echo reduction and speech quality preservation.
Solution Approach 2:
The patent implements a feedback mechanism where the correlation analysis results from comparing echo reference signals with microphone signals are used to adjust suppression parameters. This closed-loop control enables the system to automatically adapt suppression intensity based on actual signal conditions, preventing over-suppression of speech while effectively reducing residual echo, thus resolving the contradiction between echo reduction and speech fidelity.
3Measurement precision
If frequency-selective processing is implemented, then echo suppression accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent applies frequency-selective processing selectively rather than uniformly across all frequency bins. By using correlation-based detection to identify which frequency bins require suppression and applying processing only to those bins, the system achieves improved echo detection accuracy where needed while avoiding unnecessary computational overhead in bins that don't require processing, thus resolving the contradiction between precision and complexity.
Data Source
AI summary
A system configured to improve speech quality by performing residual echo suppression (RES). The system may detect when double-talk conditions are present in individual frequency bands during a voice conversation and may determine gain values for the individual frequency bands. The system may determine whether double-talk conditions are present based on a normalized cross power spectral density function in a frequency domain. If double-talk conditions are present in a frequency band or far end energy is low, the system may determine a gain value that passes audio data in the frequency band, whereas if double-talk conditions are not present, the system may determine a gain value that attenuates audio data in the frequency band. The system may determine binary gain values using a decision threshold value or continuous gain values using a mapping function. The system may control an amount of suppression by selecting different mapping functions and/or parameters.