Music Graph

Music graphs represent musical scores and audio as interconnected networks, aiming to improve music understanding and generation through graph-based algorithms and neural networks. Current research focuses on developing efficient graph processing frameworks, applying graph neural networks (GNNs) to various music tasks (e.g., melody prediction, harmonization), and exploring hierarchical models to capture long-term musical structures. This approach offers a powerful paradigm shift in symbolic music processing, enabling more sophisticated music generation, analysis, and human-computer interaction in music creation.