Harnessing Multimodal Music Hooks for Emotion-Driven Composition and Analysis

View at publisher Summarise with VIRAT
Authors: Balaji Mudadla, Vani V
Cited by: 0
References: 15

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Internet has plenty of unimodal music data, either symbolic or audio representation. However, there is a notable scarcity of multimodal music datasets that offer aligned modal in-formation and comprehensive annotations for music understanding and generation. In this paper, we introduce A multimodal music dataset called Popular Hooks is available to the public. It contains 38,694 popular musical hooks, which are distinctive parts of songs, together with synchronized MIDI, music videos, audio, and lyrics. In addition, the dataset offers comprehensive annotations of broad musical characteristics like structure, genre, mood, and location. Precisely, we use a pre-trained multimodal music emotion recognition system in automatic emotion labeling and we undertake user study to determine its accuracy. Lastly, we discuss how this dataset can be used to generate the baselines of emotion-conditioned music generation, proving that it is capable of contributing to the field.

Fields of Study & Subjects

Music and Audio Processing Music Technology and Sound Studies Emotion and Mood Recognition

Publication details

Journal / source
International Conference on Computing, Communication, Control and Cyber-Physical Systems (I5CPS)
Pages
1–6
Publication date
26 February 2026
Language
EN