A Method for Separating Music and Human Voice Based on U-shaped Network and Audio Fingerprint

A technology of audio fingerprint and separation method, which is applied in speech analysis, instruments, etc., and can solve the problem that the model cannot obtain speech

Active Publication Date: 2021-08-17
HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN (INSTITUTE OF SCIENCE AND TECHNOLOGY INNOVATION HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN)
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, in reality, due to some conditions, the model obtained simply through the speech separation algorithm will not obtain very pure speech

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
  • A Method for Separating Music and Human Voice Based on U-shaped Network and Audio Fingerprint
  • A Method for Separating Music and Human Voice Based on U-shaped Network and Audio Fingerprint
  • A Method for Separating Music and Human Voice Based on U-shaped Network and Audio Fingerprint

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0044] In order to further describe the technical solution of the present invention in detail, this embodiment is implemented on the premise of the technical solution of the present invention, and provides detailed implementation methods and specific steps.

[0045] A method for separating music and human voice based on U-shaped network and audio fingerprint provided by the present invention, the specific process is as follows figure 1 Shown:

[0046] Step 1: Input the mixed audio into the trained music-vocal separation model to obtain separated accompaniment audio;

[0047] The music vocal separation model is based on a U-shaped network architecture, such as figure 2 As shown, the U-shaped network architecture includes an Encoder Encoder, a Decoder Decoder, and an intermediate layer that plays a connecting role, wherein the number of Encoder and Decoder is 6, and the specific design of Encoder and Decoder is more flexible in the embodiment. Between Encoder and Decoder In the...

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 present invention involves a method and human voice separation method based on U -shaped networks and audio fingerprints: when entering hybrid audio, first through the trained music separation model to obtain the separation accompaniment audio; then use fingerprint extraction algorithm to extract it out to extract out the fingerprint extraction algorithm to extract it out to extract out and extract it.Fingerprints in the accompaniment audio; then send these fingerprints into the audio fingerprint database to retrieve the corresponding matching songs and time points of offset; then adjust the volume of the retrieved songs;The sampling point is matched to get a period of audio waveform; then accurately match the separation audio with the volume of this waveform; finally mixed audio minus this waveform and finally get a more pure audio.This method designs a U -shaped network architecture -based music separation model, and adding audio fingerprint retrieval after the model to separate the separation of the separation effect, which significantly reduces the gap between the separation effect and obtains pure human voice.

Description

technical field [0001] The invention belongs to the technical field of speech separation, and in particular relates to a method for separating music and human voice based on a U-shaped network and audio fingerprints. Background technique [0002] Among the many speech signal processing technologies, the more representative technologies are speech recognition, speech synthesis and speech separation. Among them, speech separation is an important research direction in the field of speech. The task of speech separation is to separate target speech from background noise. Single-channel speech separation has become a research hotspot in the field of speech separation in recent years. The traditional methods of speech separation include spectrum subtraction, adaptive filter and Wiener filter, etc. These methods are relatively simple, but these methods assume that the background noise is stable. If this assumption is not satisfied, the ideal , especially when the value of SNR is a...

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): G10L21/0272G10L21/028G10L21/0308
CPCG10L21/0272G10L21/028G10L21/0308
Inventor 汤步洲刘凯隆刘超
Owner HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN (INSTITUTE OF SCIENCE AND TECHNOLOGY INNOVATION HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN)
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap