Supercharge Your Innovation With Domain-Expert AI Agents!

A method and system for intelligent music generation based on improved neural network

A neural network and convolutional neural network technology, which is applied in the field of intelligent music generation methods and systems, can solve the problems of high time and economic costs, and difficulty in making soundtracks, so as to reduce time costs and economic costs, reduce soundtrack production costs, high degree of freedom

Active Publication Date: 2022-05-03
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] In order to solve the problems existing in the prior art, the present invention provides an intelligent music generation method and system based on the improved neural network, utilizes the characteristics of computer quickness and economy, designs a deep neural network structure and trains it, Constructed a motion-driven intelligent music generation system, realized the intelligent processing of motion video and generated original audio data, and then synthesized the function of soundtrack, which solved the difficulty, time and economic cost of soundtrack production mentioned in the above-mentioned background technology high problem

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 and system for intelligent music generation based on improved neural network
  • A method and system for intelligent music generation based on improved neural network
  • A method and system for intelligent music generation based on improved neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0042]The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

[0043] see Figure 1-6 , the present invention provides a technical solution: an intelligent music generation method based on an improved neural network, the method flow is as follows figure 1 shown, including the following steps:

[0044] Step 1: Build the dataset. First, download a large number of sports videos with soundtracks on video websites (mixed cuts with soundtracks, collections, etc.); secondly, write a python program to automatically and equal-len...

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 discloses an intelligent music generation method and system based on an improved neural network, comprising: constructing a data set, extracting initial features of video frame images, timing correlation of video frame features, generating audio original data, network training and testing. The present invention utilizes the characteristics of computer quickness and economy, designs a deep neural network structure and trains it, realizes the intelligent processing of motion video and generates original audio data, and then synthesizes the soundtrack, solving the problem of the current soundtrack production Problems of high difficulty, high time and economic cost, at the same time established a motion-driven intelligent music generation function system, which can accurately extract features from input video data, generate high-quality music and match music with sports scenes, generate The subjective score of music, MOS>3.5, realizes fast and batch soundtrack generation for videos with sports and other themes, and more than doubles the time cost and economic cost of music production.

Description

technical field [0001] The invention relates to the technical fields of computer vision and intelligent soundtracking, in particular to an intelligent music generation method and system based on an improved neural network. Background technique [0002] Exercise is a variety of activities that human beings cultivate their physical fitness in the process of development. Since the beginning of the 21st century, with the continuous advancement of science and technology, people have begun to seek to integrate various elements into sports, such as: adding music to sports videos to give the audience a stronger audio-visual experience. Music is an important form of expression in the field of art, which embodies the human thinking mode and is a unified combination of regularity and creativity. On the one hand, the production of music is based on certain knowledge of music theory; on the other hand, music itself carries certain human emotions. [0003] The technology of using deep l...

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): H04N21/439H04N21/44H04N21/81G06N3/04G06N3/08
CPCH04N21/44008H04N21/4394H04N21/44012H04N21/4398H04N21/8113H04N21/8166G06N3/084G06N3/044G06N3/045
Inventor 吴庆波施兆丰李宏亮孟凡满许林峰潘力立
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More