[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Music, emotion, memory of passages and content analysis (LSA)



Hi, Daniel:
 
I am a phd student of EECS.  I don't know if it is relevant, but I have some preliminary results related to the automatic prediction of emotion values of music.
 
http://mpac.ee.ntu.edu.tw/~yihsuan/publication.html
 
Y.-H. Yang et al, "A regression approach to music emotion recognition," IEEE Transactions on Audio, Speech and Language Processing (TASLP), vol. 16, no. 2, pp. 448-457, Feb. 2008.
 
In this paper, we formulate music emotion recognition as a regression problem and predict the arousal and valence values (numerical values) of music.  The ground truth data needed for training an automatic regression model is obtained through a subjective test.  Subjects are asked to rate the arousal and valence values of a number of songs.  Features are extracted from the audio signal to represent the songs, and support vector regression (SVR) is adopted to train the regression model.

Y.-H. Yang and H.-H. Chen, "Music emotion ranking," in Proc. IEEE Int. Conf. Acoustics, Speech, and Signal Processing 2009 (ICASSP'09), Taipei, Taiwan, accepted.

The cognitive load of rating emotion may be too high.  In this paper, we propose a ranking measure and ask subject to annotate emotion in a comparative way.
 
 
 

Sincerely yours,

--
Yi-Hsuan Yang (Eric), Ph.D. candidate,
MPAC Lab,
Graduate Institute of Communication Engineering,
National Taiwan University.
http://mpac.ee.ntu.edu.tw/~yihsuan/