EmotionBox: A music-element-driven emotional music generation system based on music psychology
Kaitong Zheng, Ruijie Meng, Chengshi Zheng, Xiaodong Li, Jinqiu Sang, Juanjuan Cai, Jie Wang, Xiao Wang
Abstract
Open-access reader
Kaitong Zheng, Ruijie Meng, Chengshi Zheng, Xiaodong Li, Jinqiu Sang, Juanjuan Cai, Jie Wang, Xiao Wang
Abstract
Open-access reader
With the development of deep neural networks, automatic music composition has made great progress. Although emotional music can evoke listeners' different auditory perceptions, only few research studies have focused on generating emotional music. This paper presents EmotionBox -a music-element-driven emotional music generator based on music psychology that is capable of composing music given a specific emotion, while this model does not require a music dataset labeled with emotions as previous methods. In this work, pitch histogram and note density are extracted as features that represent mode and tempo, respectively, to control music emotions. The specific emotions are mapped from these features through Russell's psychology model. The subjective listening tests show that the Emotionbox has a competitive performance in generating different emotional music and significantly better performance in generating music with low arousal emotions, especially peaceful emotion, compared with the emotion-label-based method.
OpenAlex reports 20 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
With the development of deep neural networks, automatic music composition has made great progress. Although emotional music can evoke listeners' different auditory perceptions, only few research studies have focused on generating emotional music. This paper presents EmotionBox -a music-element-driven emotional music generator based on music psychology that is capable of composing music given a specific emotion, while this model does not require a music dataset labeled with emotions as previous methods. In this work, pitch histogram and note density are extracted as features that represent mode and tempo, respectively, to control music emotions. The specific emotions are mapped from these features through Russell's psychology model. The subjective listening tests show that the Emotionbox has a competitive performance in generating different emotional music and significantly better performance in generating music with low arousal emotions, especially peaceful emotion, compared with the emotion-label-based method.
Key concepts: Music and emotion, Music psychology, Psychology, Cognitive psychology, Arousal, Active listening, Musicality, Perception