Residual Echo Filtering Using Wiener Spectral Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing echo cancellation methods, such as adaptive filters, struggle to completely eliminate residual echo due to factors like background noise, non-linear distortions, and changes in the echo path, which affects user experience and speech recognition rates.
Innovation Solution
A method that performs echo cancellation by using an echo reference signal to estimate residual echo, and then applies a Wiener filter coefficient determined by the a priori signal-to-noise ratio to filter out the residual echo, effectively reducing voice pollution and improving voice quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If an adaptive filter is used for echo cancellation, then the echo signal can be generated and subtracted to achieve echo cancellation, but residual echo remains due to background noise, non-linear distortion, speaker delay, and sudden changes in the echo path
Solution Approach 1:
The patent introduces a spectral subtraction mechanism as an intermediary process between echo cancellation and voice output. The spectral subtractor acts as a mediator that identifies and removes residual echo components from the voice signal spectrum, thereby eliminating the harmful residual echo that the adaptive filter cannot completely eliminate
Solution Approach 2:
The patent replaces the traditional time-domain adaptive filtering approach with a frequency-domain spectral analysis approach. By transforming the voice signal to the frequency domain and performing spectral subtraction, the system achieves more effective residual echo removal compared to conventional mechanical filtering methods
2Measurement precision
If spectral subtraction is performed using a priori signal-to-noise ratio, then residual echo can be accurately filtered, but the calculation complexity increases
Solution Approach 1:
The patent performs preliminary estimation of the a priori signal-to-noise ratio using the echo reference signal before the actual spectral subtraction process. This preliminary action allows the system to prepare the necessary parameters for accurate residual echo filtering in advance, improving filtering accuracy while managing computational complexity through staged processing
Data Source
AI summary
Disclosed is a method for reducing residual echo including: performing an echo cancellation process on a voice input signal according to an echo reference signal to obtain an echo cancellation signal; performing a FFT on the echo reference signal to obtain a reference spectrum signal for each frame; performing the FFT on the echo cancellation signal to obtain a speech spectrum signal for each frame; using the reference spectrum signal and the speech spectrum signal of a current frame to obtain a priori signal-to-noise ratio of the current frame according to a principle of additive noise; filtering the speech spectrum signal of the current frame by a Wiener filter coefficient of the current frame determined by the priori signal-to-noise ratio of the current frame to obtain a target spectrum signal of each frame; performing an IFFT on the target spectrum signal of each frame to obtain a target voice signal.