Method and system for judging double-end conversation and method and system for eliminating echo

Abstract

The invention provides a method and a system for eliminating echo in a communication system with a loudspeaker and a microphone, the method comprises the following steps: carry out the noise energy estimation processing to a near-end microphone signal to obtain an estimated noise signal, wherein, the near-end microphone signal comprises a possible near-end speech signal, a possible background noise signal, an echo signal which is formed by outputting a possible far-end speech signal from the loudspeaker and being received by the microphone; carry out the echo estimation processing to the far-end speech signal to obtain an estimated echo signal, wherein, the far-end speech signal is output from the loudspeaker; obtain an error signal by the difference of the near-end microphone signal and the estimated echo signal; calculate the variance of the error signal and the variance of the estimated noise signal; calculate an judging factor, wherein, the judging factor corresponds to the variance of the error signal and the variance of the estimated noise signal; compare the judging factor and a predetermined threshold value, when the judging factor is less than the predetermined threshold value, judge a conversation which is not a double-end conversation, otherwise, judge the conversation which is the double-end conversation.

Description

technical field [0001] The present invention relates to a method and system for eliminating echo, more particularly, to a method and system for judging double-end dialogue in a communication system. Background technique [0002] Hands-free communication systems with microphones and speakers, which can be used for mobile phones, conference calls, etc., are now becoming more and more popular. During hands-free communication, at least two users engage in a conversation using the hands-free communication system. Usually, the voice of one of the at least two users is called the near-end voice, and the voice of the other user is called the far-end voice. Since the far-end voice signal is output through the near-end speaker, partially input into the near-end microphone, and incorrectly transmitted to other users at the far end, the echo of the far-end voice is caused, thus causing the user to hear himself or echoes of other people's speech. In this case, the echo should be remov...

AI Technical Summary

Problems solved by technology

However, in this method, in practical situations, it is difficult to choose the appropriate thresholds T1 and T2 since different transition times may pass from changing from double-ended to single-ended or from single-ended to double-ended to judge double-ended conversations as accurately as possible

In addition, since the double-talk judging device does not consider the influence of noise and echo path nonlinearity, the performance of the double-talk judging device will degrade when the speaker-microphone environment contains noise and nonlinearity of the echo path

Images