Appliance Of Deep Learning In Music Information Retrieval

In recent years,the development of Deep Learning technology,with its various excellent application results in scientific research and industry,widely appears in the vision of the public and professional practitioners.In the field of art,Deep Learning technology also has some significant applications,such as artificial intelligence composition,auxiliary composition and mixing,etc.Through the construction and development of Deep Neural Network,it has brought fresh blood to the future world for art.The purpose of this paper is to explore how Deep Learning can be applied in the Music Information Retrieval,which will bring new technical methods to the research field and develop a more intelligent working mode.Combing the current feasible audio processing methods and deep neural network model,and using Tensorflow 2.0 as the platform,the author designed and built a Deep Neural Network to solve a specific MIR task-"music genre classification".After getting the expected practical results,it obviously shows the feasibility and advantages of the application of Deep Learning technology in MIR.In addition,the prospect of its possible technology derivatives and its application in the future is discussed.