Generative AI for Music and Audio

Generative AI has been transforming the way we interact with technology and consume content. In the next decade, AI technology will reshape how we create audio content in various media, including music, theater, films, games, podcasts, and short videos. In this dissertation, I introduce the three main directions of my research centered around generative AI for music and audio: 1) multitrack music generation, 2) assistive music creation tools, and 3) multimodal learning for audio and music. Through my research, I aim to answer the following two fundamental questions: 1) How can AI help professionals or amateurs create music and audio content? 2) Can AI learn to create music in a way similar to how humans learn music? My long-term goal is to lower the barrier of entry for music composition and democratize audio content creation