Voice endpoint determination method and device, storage medium and electronic device

A technology to determine the method and voice, applied in voice analysis, voice recognition, instruments, etc., can solve the problems of low accuracy rate, achieve the effect of improving recognition accuracy and solving the effect of low accuracy rate
CN110706693AActive Publication Date: 2020-01-17ZHEJIANG DAHUA TECH CO LTD
7 Cites 9 Cited by

Patent Information

Authority / Receiving Office
CN ยท China
Patent Type
Applications(China)
Current Assignee / Owner
ZHEJIANG DAHUA TECH CO LTD
Publication Date
2020-01-17

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The embodiment of the invention provides a voice endpoint determination method and device, a storage medium and an electronic device, and the method comprises the following steps: carrying out the preprocessing of an obtained audio signal, obtaining a plurality of sub-bands, enabling the audio signal to comprise N audio signal frames, enabling N to be an integer greater than 1, and enabling the sub-bands to be obtained through the division of the audio signal frames based on a frequency band; obtaining the ratio of the signal-to-noise ratio to the spectral entropy of the audio signal frame according to the ratio of the signal-to-noise ratio to the spectral entropy of the sub-band; judging whether the audio signal frame is a voice frame or not by using a double-threshold detection algorithmaccording to the ratio of the signal-to-noise ratio to the spectral entropy of the audio signal frame; and if so, respectively determining the first voice frame and the last voice frame of the audiosignal as a voice starting endpoint and a voice ending endpoint of the audio signal. The problem of low accuracy due to the fact that voice endpoint detection only aims at a certain single feature inthe prior art is solved.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The present invention relates to the field of audio, video and communication technologies, in particular, to a method and device for determining a voice endpoint, a storage medium, and an electronic device. Background technique

[0002] Speech endpoint detection is an important link in the field of speech information processing. In many practical applications such as speech response systems, speaker recognition systems and speech recognition systems, it is required to first judge the input signal of the system and accurately find out the location of the speech segment. The start point and the end point, so that the collected data can really be a valid voice signal, which can reduce the amount of transmitted data and calculation and reduce the processing time. In the current related technology, the voice endpoint detection only detects a single feature, and the accuracy rate is low.

[0003] Aiming at the problem in related technologies that the voice ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More