• Version
• Download 8
• File Size 630.10 KB
• File Count 1
• Create Date 28 April 2023
• Last Updated 28 April 2023

A MUSIC RECOMMADATION BASED ON EMOTION DETECTION USING DEEP CONVOLOTION NEURAL NETWORK

MR.A.PANDIAN , SURYA K, SURIYAPRAKASH C, THINAGARAN T, VIGNESH S

ABSTRACT:

Human emotions are inconsistent and are truly a product of both internal and exterior events taking place in a person's environment. Human emotions have been the subject of extensive study and investment, which has opened the door to many potential uses. The current method includes automatically generating a playlist of music based on genres, artists, etc. Manually organizing audio files into playlists is still another choice. Calculating music similarity and multiple frequency estimates are recent problems. A QBSH (Query by singing and humming) method uses the song's content to determine what it is (tune and rhythm). Nevertheless, the problem with this approach is that it takes time and doesn't always satisfy the consumer. The user's emotion is not taken into account in the current system. A music recommendation system that considers human emotions can be created because they are important in daily activities. A person's emotion can be determined to determine the type of music that would work best. The technology seeks to analyses the data supplied by determining the user's emotion. The classification of the various emotions   using a deep learning algorithm is followed by the generation of labels and the playing of appropriate music. The proposed system has produced results with a notable degree of accuracy and it also opens the door for additional study in this field.

Keywords:  QBSH, Emotions, Deep learning