Cross-Dataset and Cross-Cultural Music Mood Prediction: A Case on Western and Chinese Pop Songs
Authors:
- 20 May 2017
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Affective Computing
- Vol. 8 (2) , 228-240
Abstract
In music mood prediction, regression models are built to predict values on several mood-representing dimensions such as valence (level of pleasure) and arousal (level of energy). Many studies have shown that music mood is generally predictable based on music acoustic features, but these experiments were mostly conducted on datasets with homogeneous music. Little research has been done to explore the generalizability of mood regression models cross datasets, especially those with music in different cultures. In the increasingly global market of music listening, generalizable models are highly desirable for automated processing, searching and managing music collections with heterogeneous characteristics. In this study, we evaluated mood regression models built on fifteen acoustic features in five mood-related musical aspects, with a focus on cross-dataset generalizability. Specifically, three distinct datasets were involved in a series of five experiments to examine the effects of dataset size, reliability of annotations and cultural backgrounds of music and annotators on mood regression performances and model generalizability. The results reveal that the size of the training dataset and the annotation reliability of the testing dataset affect mood regression performances. When both factors are controlled, regression models are generalizable between datasets sharing a common cultural background of music or annotators.Keywords
Funding Information
- University of Hong Kong (NSC 102-2221-E-001-004-MY3)
- Ministry of Science and Technology of Taiwan
This publication has 26 references indexed in Scilit:
- The AMG1608 dataset for music emotion recognition2015
- Investigating the Impact of Sound Angular Position on the Listener Affective StateIEEE Transactions on Affective Computing, 2015
- Predictive Modeling of Expressed Emotions in Music Using Pairwise ComparisonsLecture Notes in Computer Science, 2013
- A Review of Music and Emotion Studies: Approaches, Emotion Models, and StimuliMusic Perception: An Interdisciplinary Journal, 2012
- Automatic recognition of emotion evoked by general sound events2012
- What does music mood mean for real users?2012
- Musical emotions: Predicting second-by-second subjective feelings of emotion from low-level psychoacoustic features and physiological measurements.Emotion, 2011
- Extracting Predominant Local Pulse Information From Music RecordingsIEEE Transactions on Audio, Speech, and Language Processing, 2010
- Automatic mood detection and tracking of music audio signalsIEEE Transactions on Audio, Speech, and Language Processing, 2005
- Cue utilization in communication of emotion in music performance: Relating performance to perception.Journal of Experimental Psychology: Human Perception and Performance, 2000