Music Emotion Detection: Transfer Learning Approach

Conference: Second Joint International Conferences on Control System and Information Technology
Author(s): Pallavi Jaiswal, Bhavana Tiple Year: 2018
Grenze ID: 02.CSIT.2018.2.505 Page: 13-17


Music plays an important role in day to day life. It has power to stimulate various\nemotions within us. In the field of psychological studies emotion detection has become\nimportant research topic. Emotion detection in music will support bonding between human\nand music. This will not only help to understand psychological state of mind but also\nprovide improved music information retrieval system. We have selected term Raaga as the\nbasic model of our work were different emotions are associated different Raaga. In current\nwork we are using Indian Classical music database. In the proposed system, music features\nare selected and extract to map the relationship between music and emotion. Transfer\nlearning is the upgrading version of machine learning. Were our current task is formed by\nusing knowledge learned while implanting related domain task. Thus, in the proposed\nsystem we will be using transfer learning method to map various emotions from music.


CSIT - 2018