Prediction of the Distribution of Perceived Music Emotions Using Discrete Samples

• Published in: IEEE transactions on audio, speech, and language processing Vol. 19; no. 7; pp. 2184 - 2196
• Main Authors: Yi-Hsuan Yang, Chen, H. H.
• Format: Journal Article
• Language: English
• Published: Piscataway, NJ IEEE 01.09.2011; Institute of Electrical and Electronics Engineers
• Subjects: Accuracy; Applied sciences; Arousal; Clips; Computational modeling; emotion distribution prediction; Emotions; Exact sciences and technology; Feature extraction; Information, signal and communications theory; Kernel; Machine learning; Mathematical models; Miscellaneous; Music; music emotion recognition; Planes; Predictive models; regression; Samples; Signal processing; Statistical analysis; Statistical methods; subjectivity; Telecommunications and information theory; Training; valence; Performance evaluation; emotion distribution prediction; Emotion recognition; subjectivity; Probabilistic approach; Probability distribution; valence; Algorithm; Modeling; Sound recognition; Learning; Calling line identification presentation; regression; Arousal; Automatic recognition; Musical sound; music emotion recognition
• ISSN: 1558-7916; 1558-7924
• DOI: 10.1109/TASL.2011.2118752

Typically, a machine learning model of automatic music emotion recognition is trained to learn the relationship between music features and perceived emotion values. However, simply assigning an emotion value to a clip in the training phase does not work well because the perceived emotion of a clip varies from person to person. To resolve this problem, we propose a novel approach that represents the perceived emotion of a clip as a probability distribution in the emotion plane. In addition, we develop a methodology that predicts the emotion distribution of a clip by estimating the emotion mass at discrete samples of the emotion plane. We also develop model fusion algorithms to integrate different perceptual dimensions of music listening and to enhance the modeling of emotion perception. The effectiveness of the proposed approach is validated through an extensive performance study. An average R 2 statistics of 0.5439 for emotion prediction is achieved. We also show how this approach can be applied to enhance our understanding of music emotion.