Intelligent Mining and Retrieval of Music Data Based on Long-Term and Short-Term Memory Neural Network
Gao Yan, Lin Gao · 2023
In the era of digitalization and networking, multimedia content represented by images, audio and video has become the most important data carrier on the Internet Information superhighway. The traditional single word search method can no longer meet the needs of digital multimedia information search. At the same time, content-based multimedia information search technology has put forward higher requirements for devices and users. Therefore, search technology based on music emotions has emerged. However, currently, most music searches are based on music style, song titles, album titles, etc., without considering the emotional factors of the music itself. So, searching by “music emotion category” is currently a requirement for applications. This project aims to explore the functional relationship between music feature information and music emotion cognition from the perspective of machine learning, and compare the performance of SVM algorithms; Based on the emotional database of Chinese ethnic music, BP neural network, linear regression and other methods are used to model and optimize the emotions of music, ultimately achieving accurate recognition of music emotions.