Combined de-reverberation echo cancellation method and device

Abstract

The invention relates to a combined de-reverberation echo cancellation method, which comprises the following steps: S1, acquiring an analog microphone signal and an analog reference signal by using a microphone array and an audio playing device respectively, and converting the analog microphone signal and the analog reference signal into a digital microphone signal and a digital reference signal through an ADC respectively; S2, decomposing the received time domain signal of each sound channel of the digital microphone signal and the digital reference signal into frequency domain signals of a plurality of frequency points through short-time Fourier transform to obtain frequency domain microphone signals and frequency domain reference signals; S3, taking the frequency domain microphone signals and the frequency domain reference signals of the current frame and a plurality of past frames of the current frame as cached frequency domain signals and storing the cached frequency domain signals in a joint prediction buffer; and S4, filtering the frequency domain signals cached in the joint prediction buffer through an adaptive filter, and carrying out echo cancellation on the frequency domain microphone signals in the frequency domain signals. According to the scheme of the invention, an echo cancellation algorithm combined with dereverberation is utilized, the echo cancellation effect can be effectively improved, and the voice interaction quality is improved.

Description

technical field [0001] The invention belongs to the technical field of speech recognition, and in particular relates to an echo cancellation method and device for joint de-reverberation. Background technique [0002] In recent years, artificial intelligence technology and the Internet of Things have developed rapidly and landed. Speech recognition, as an important means of human-computer interaction, plays a key role in it. However, in the actual landing process, complex acoustic application environments, harsh user Experience limits the effective promotion of products, especially for audio systems with speakers and microphones. The quality of echo cancellation directly affects the user's voice recognition interaction experience. Therefore, how to effectively improve the echo cancellation effect is a key issue in improving the quality of speech recognition interaction. [0003] At present, the commonly used echo cancellation method estimates the echo channel through differe...

AI Technical Summary

Problems solved by technology

[0003] At present, the commonly used echo cancellation method estimates the echo channel through different adaptive filters, and then eliminates the echo. This type of method has certain effectiveness, but it ignores the reverberation in the environment, which often affects the self in the actual application process. Estimated effect of adaptive filter

At the same time, there is another type of echo cancellation method based on deep neural network. This type of method has obvious effects on solving the problem of nonlinear distortion in the echo cancellation process, but the training samples required by this type of method are relatively harsh, and its actual In the process of landing, factors such as computing power and memory of the product restrict its development

Images