Unlock instant, AI-driven research and patent intelligence for your innovation.

An Audio Sentence Boundary Detection Method

A boundary detection and sentence technology, applied in speech analysis, speech recognition, instruments, etc., to achieve the effect of flexibility, simple method and strong practicability

Inactive Publication Date: 2021-07-13
FUZHOU UNIV
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, there are very few studies related to singing audio sentence segmentation, and some of the work is essentially aimed at the segmentation of text sentences.

Method used

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
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • An Audio Sentence Boundary Detection Method
  • An Audio Sentence Boundary Detection Method
  • An Audio Sentence Boundary Detection Method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0033] The technical solution of the present invention will be specifically described below in conjunction with the accompanying drawings.

[0034] The present invention proposes an audio sentence boundary detection method, such as figure 1 The shown is mainly divided into three steps: firstly, use the improved spectral entropy algorithm based on CQT to perform routine endpoint detection on the singing signal, then analyze and classify the sentence endpoints in combination with the K-Means algorithm, and finally analyze the sentence boundary sequence to the singing signal to split. details as follows:

[0035] Step S1: Calculating the CQT spectrum of the singing voice audio: First, filter the entire singing voice audio signal, pre-emphasize, normalize and other preprocessing, and then perform CQT transformation on the pre-processed singing voice x(n) to obtain the frequency domain signal X CQT (k):

[0036]

[0037] in, is the window length N k Hanning window function...

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

PUM

No PUM Login to View More

Abstract

The invention relates to an audio sentence boundary detection method. Aiming at the sentence segmentation problem of singing voice, according to the characteristics of human auditory perception, CQT is used to perform time-frequency conversion on singing voice, and fine-grained endpoint detection is performed on singing voice through sub-band spectral entropy method; Due to the characteristics of long pauses and long pronunciation intervals, the position of the endpoints of the clauses is obtained by clustering analysis based on the K-means algorithm, which can better obtain the boundary points between sentences in the a cappella audio, and adapt to music of different rhythms. An audio sentence boundary detection method proposed by the present invention is simple, flexible to realize and has strong practicability.

Description

technical field [0001] The invention relates to the field of audio segmentation, in particular to an audio sentence boundary detection method. Background technique [0002] Audio segmentation is a crucial processing technique in fields such as music retrieval, speech recognition, and audio scene analysis. In the face of tens of thousands of audio data in the current big data era, if the audio data stream can be automatically segmented according to different contents, it will greatly simplify manual operations and improve efficiency. However, there are very few studies related to singing audio sentence segmentation, and some of the work is essentially aimed at the segmentation of text sentences. Therefore, it is of great practical application value to study how to perform sentence boundary detection on a cappella audio. Contents of the invention [0003] The purpose of the present invention is to provide an audio sentence boundary detection method, which can analyze the b...

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

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G10L15/04G10L15/08G10L25/18G10L25/45G06K9/62
CPCG10L15/04G10L15/08G10L25/18G10L25/45G06F18/23213
Inventor 张栋彭建云肖小强
Owner FUZHOU UNIV