A voice signal positioning method, device and system based on complex environment

Abstract

The embodiments of the invention disclose a complex environment-based voice signal positioning method, device and system. The method includes the following steps that: echo cancellation processing isperformed on at least two paths of expected signals according to loudspeaker reference signals, so that at least two paths of first residual signals are acquired; echo suppression processing is performed on each path of first residual signals of the at least two paths of first residual signals according to a preset echo suppression transfer function, so that at least two paths of second residual signals can be obtained; noise suppression processing is performed on each path of second residual signals of the at least two paths of second residual signals according to a preset noise suppression transfer function, so that at least two paths of third residual signals are obtained; and when it is determined that ambient sound is in a voice state at present according to the at least two paths ofthird residual signals, voice signals in an environment are located. In this way, the interference of non-target speech state signals on sound source positioning is reduced, the accurate positioning of a sound source is realized, and the robustness of the positioning of the sound source is improved.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of signal processing, and in particular to a voice signal positioning method, device and system based on complex environments. Background technique [0002] The sound source localization algorithm is generally based on an array composed of multiple microphones, and uses the phase difference between multiple input audio signals to estimate the sound source direction information. However, when the speaker of the audio device plays the audio signal and is in a high-noise environment, there are high-energy echo signals and noise signals in the microphone signal, and these signals will have a strong impact on the location of the target voice source by the sound source localization algorithm. Interference, resulting in a large deviation in the estimation of the orientation information of the algorithm and low robustness. [0003] In order to solve the above problems, algorithms such as echo ...

AI Technical Summary

Problems solved by technology

[0005] For this reason, the embodiment of the present invention provides a voice signal localization method, device and system based on a complex environment, to solve the problem of the sound source localization in the prior art when in a complex environment with high noise and strong nonlinear echo. Accurate, low robustness issues

Images