Music Transformer

Music Transformers are deep learning models leveraging the transformer architecture to generate, analyze, and manipulate music. Current research focuses on improving music generation quality and controllability through novel architectures like nested transformers and the incorporation of additional information such as chords, instruments, and even video content to condition the generation process. These advancements aim to create more expressive and musically coherent AI-generated music, impacting both music composition and applications like optical music recognition (OMR) and video soundtrack generation. The development of large, pre-trained music transformers and improved data representations are also key areas of ongoing investigation.