MICW: A Multi-Instrument Music Generation Model Based on the Improved Compound Word
Published in IEEE International Conference on Multimedia and Expo Workshops (ICMEW), 2022
In this work, we address the task of multi-instrument music generation. Notably, along with the development of artificial neural networks, deep learning has become a leading technique to accelerate the automatic music generation and is featured in many previous papers like MuseGan, MusicBert, and PopMAG. However, seldom of them implement a well-designed representation of multi-instrumental music, and no model perfectly introduces a prior knowledge of music theory. In this paper, we leverage the Compound Word and R-drop method to work on multi-instrument music generation tasks. Objective and subjective evaluations show that the generated music has cost less training time, and achieved prominent music quality.
@INPROCEEDINGS{9859531,
author={Liao, Yikai and Yue, Wang and Jian, Yuqing and Wang, Zijun and Gao, Yuchong and Lu, Chenhao},
booktitle={2022 IEEE International Conference on Multimedia and Expo Workshops (ICMEW)},
title={MICW: A Multi-Instrument Music Generation Model Based on the Improved Compound Word},
year={2022},
volume={},
number={},
pages={1-10},
keywords={Training;Deep learning;Costs;Conferences;Artificial neural networks;Compounds;Task analysis;Multi-Instrument;R-Drop},
doi={10.1109/ICMEW56448.2022.9859531}}
Recommended citation: Y. Liao, W. Yue, Y. Jian, Z. Wang, Y. Gao and C. Lu, "MICW: A Multi-Instrument Music Generation Model Based on the Improved Compound Word," 2022 IEEE International Conference on Multimedia and Expo Workshops (ICMEW), Taipei City, Taiwan, 2022, pp. 1-10, doi: 10.1109/ICMEW56448.2022.9859531.
Download Paper