About

My research connects music information retrieval, generative modeling, and musical expression. I work on representations that connect scores with performances, from piano performance modeling to ornamentation and heterophonic ensemble generation.

I studied Music at Quanzhou Normal University (MA, September 2023–June 2026), supervised by Prof. Jianbing Xiahou. My dissertation, Deep Learning Based Music Generation on Nanyin Musical Features, examined AI-based generation for this musical tradition. My earlier training was in Communication Engineering at China West Normal University.

My collaborative work includes MAJEPPA, a framework for piano performance representation, and NanyinHGNN, which uses a pipa-centric dataset and heterogeneous graphs for Nanyin generation. My individual projects explore expressive MIDI ornamentation, performance rendering, dynamic audio tokenization, and failures in latent-space interpolation.

Alongside these studies, I develop interactive performance systems that use music analysis as a compositional mechanism. I am interested in how representation learning and interactive interfaces can support creative exploration while attending to the musical practices of performers and traditions.

Education & background