Research on music score recognition based on deep confidence network
How to effectively manage and identify music element information is the focus of digital music re-search in recent years.Therefore,based on the deep confidence network,a music score recognition algo-rithm is proposed to extract music feature information from multiple perspectives,analyze from the perspec-tive of music emotion,optimize the deep confidence model,and finally input the extracted music feature da-ta into the optimized deep confidence model to complete the recognition and classification of music.Through experimental analysis,in the test of different music score tags,compared with other recognition models,the optimized deep confidence recognition model can more accurately identify each emotion tag,with an average of 91.14％.The recognition performance is the better,meeting the recognition and management require-ments of music score.The research content can provide technical reference for the retrieval and management of music data.
deep confidencemusic recognitionemotion classificationmultiple featuresmusic label
NETL
NSTL
万方数据
