Design Of Intelligent VLOG Generation System Based On The Fusion Of Music And Scene

With the new media gradually showing the development trend of technological empowerment,video blog(Vlog)is presented to the public in an active growth manner,and it has become a popular sharing form with sophisticated and high-quality contents.In this environment,video shooting and editing softwares of Vlog are iteratively updated,and the pain points such as homogeneity of software functions,lack of innovation and intelligence also appear.Therefore,this paper aims to provide better and more reasonable theoretical support and solutions for Vlog creation,and provide smarter and more convenient operation guides and tool platforms for video creators,and designs an intelligent Vlog generation system based on tne fusion of music and scene.The main work of this paper for intelligent Vlog generation is as follows:(1)Firstly,MES model,a design model of intelligent Vlog generation system based on the fusion of music and scene,is proposed.The model contains three key elements: music,emotion and scene,and the three are related to each other.This model supports the design methods of intelligent Vlog generation system based on user,data and technology,which lays a theoretical foundation for the work of this paper.(2)A scene library containing 6 major Vlog popular scenes and a Vlog music library were constructed,and the mapping relationship between ”music-emotion”,”music-scene”and ”emotion-scene” was carried out based on the MES model as the research theoretical basis.The experimental results show that: absolute music emotions are affected by scene emotions,and the correct combination of scenes and music can greatly improve the overall audiovisual experience of the video,and the information and expressed emotions of the music itself will also affect the user's perception of the scene select;(3)A method of music analysis and video synthesis based on music rhythm recognition is proposed,which has the feature of automatically marking music rhythm points.Through the subjective user evaluation experiment,the method in this paper is compared with the others to verify its effectiveness;(4)By integrating the above works,the AutoVlog system of web version is designed and developed.And through the user experience experiment,the system is verified to have good effectiveness,efficiency and usability.The system provides users with a personalized,intelligent and integrated creative experience,which solves the problems of homogenization,non-intelligence and complicated operation of existing video editing tools.This paper provides a useful reference for the study of the relationship among music,emotion and scene,and also contributes an application case for the video generation industry.