Deep Learning-based Intelligent Music Composition System: Assisting Composition and Arrangement

Music is a complex sound art, which is not only closely integrated with human hearing, emotion, behavior and other factors but also interconnected with many other factors such as the musical environment, the composer, and the background of the piece's creation. In practical application, the lack of methods for analyzing and summarizing musical works makes it difficult for music creators and arrangers to fully understand their creative intentions, and thus problems such as inaccurate compositions and unreasonable arrangements often occur in the actual creative process. In this paper, a new intelligent music composition system is proposed by combining the information fusion method of deep learning and the information composition method of music works. The system greatly improves the creation efficiency and provides music creators with inspiration and innovative ideas.