Human-centric music medical therapy exploration system

Authors:
Baixi Xing;Kejun Zhang;Lekai Zhang;Eng Keong Lua;Shouqian Sun
Affiliations:
Zhejiang University, Hangzhou, China;Zhejiang University, Hangzhou, China;Zhejiang University, Hangzhou, China;Monash University, Melbourne, Australia;Zhejiang University, Hangzhou, China
Venue:
Proceedings of the 2013 ACM SIGCOMM workshop on Future human-centric multimedia networking
Year:
2013

Citing 4
Cited 0

Music emotion classification: a fuzzy approach

MULTIMEDIA '06 Proceedings of the 14th annual ACM international conference on Multimedia
EEG-based emotion recognition in music listening: A comparison of schemes for multiclass support vector machine

ICASSP '09 Proceedings of the 2009 IEEE International Conference on Acoustics, Speech and Signal Processing
Feature selection for content-based, time-varying musical emotion regression

Proceedings of the international conference on Multimedia information retrieval
Prediction of the Distribution of Perceived Music Emotions Using Discrete Samples

IEEE Transactions on Audio, Speech, and Language Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Music emotion analytic is useful for many human-centric applications such as medical intervention. Existing studies have shown that music is a low risk, adjunctive and therapeutic medical intervention. However, there is little existing research about the types of music with identified emotions that have the most effect for different medical applications. We would like to discover various music emotions through machine learning analytic so as to identify modelsof how music conveys emotion features, and determine its effectiveness for medical intervention and treatment. We are developing a Human-centric Music Medical Therapy Exploration System which could recognize music emotion features from Chinese Folk Music Library, and recommend suitable music to playback for medical intervention and treatment. Our networked system is based on Support Vector Machine(SVM) algorithm to construct the models for music emotion recognition and information retrieval. Our main contributions are as follows: Firstly, we built up the Chinese folk music emotions and features library; secondly, we conducted evaluation and comparison with other algorithms such as Back Propagation(BP) and Linear Regression to set up the model construction for music emotion recognition and proved that SVM has the best performance; lastly, we integrated blood pressure and heartbeat data analytic into our system to visualize the emotion fluctuation in different music affection and make recommendation for suitable humancentric music medical therapy for hypertensive patients.