Echo elimination method based on deep learning
A technology of echo cancellation and deep learning, which is applied in neural learning methods, sound-generating devices, two-way sound reinforcement telephone systems, etc., and can solve problems such as poor echo cancellation effect and the need for post-filters
Patent Information
- Authority / Receiving Office
- CN · China
- Current Assignee / Owner
- Publication Date
- 2019-06-04
Smart Images

Figure 1 
Figure 2 
Figure 3
Abstract
Description
Technical field
[0001] The present disclosure relates to the field of computer application technology, and in particular to an echo cancellation method, device, electronic equipment, and storage medium based on deep learning. Background technique
[0002] In a communication system, when a speaker and a microphone are coupled, the microphone will pick up the signal from the speaker and its reverberation, thereby generating echo. For example, conference calls, hands-free phones and mobile communications all suffer from echo.
[0003] Echo cancellation faces many problems, such as dual talk, background noise, and nonlinear distortion. First of all, dual talk is a typical way of dialogue in communication systems, and speakers at both ends sometimes speak at the same time. However, near-end speech signals will seriously affect the convergence of the adaptive algorithm and may cause them to diverge. In addition, the signal received at the microphone contains not only echo and near-end...
Examples
Embodiment Construction
[0060] Here, an exemplary embodiment will be described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present invention. On the contrary, they are only examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.
[0061] figure 1 It is a flow chart showing a method for echo cancellation based on deep learning according to an exemplary embodiment. This deep learning-based echo cancellation method can be used in electronic devices such as smart phones and computers. Such as figure 1 As shown, the deep learning-based echo cancellation method may include step S110, step S120, step S130, ...