About

Github | LinkedIn | Google Scholar

Introduction

I am currently a postdoctoral researcher at the Center for Artificial Intelligence and Data Science (CAIDAS), University of Würzburg, Germany. I am affiliated with both the Computational Humanities (CH) group led by Prof. Christof Weiß and the Natural Language Processing (WüNLP) group led by Prof. Dr. Goran Glavaš.

My research lies at the intersection of AI and music. I am broadly interested in building AI systems that can understand and represent music across modalities, including audio, symbolic notation, and language. In particular, I am interested in multimodal learning, large scale representation learning, and foundation models for music understanding. Alongside my research, I am also an active learner of piano performance, which continues to shape my perspective on music understanding, perception, and expression.

Over the past years, my primary research focus has been automatic music transcription, which aims to convert music audio recordings into symbolic notation. More recently, I have been exploring audio-language models, multimodal learning, and large-scale pretraining approaches for music-related tasks, including music transcription, beat tracking, and performance analysis.

Before joining CAIDAS, I completed my PhD at the UKRI Centre for Doctoral Training in AI and Music (AIM CDT) at Centre for Digital Music (C4DM), Queen Mary University of London (QMUL), supervised by Dr. Emmanouil Benetos, Dr. Veronica Morfi and Prof. Simon Dixon.

Beyond research, I am one of the organizers of the WIMIR Mentoring Programme, an initiative supporting students from underrepresented groups in the ISMIR community by connecting junior researchers with senior mentors.

Contact

Universität Würzburg
Centre for Philology and Digitality (ZPD)
Hubland Nord
Emil-Hilb-Weg 23
97074 Würzburg
Germany

Office: 02.007, ZPD (alternatively at 02.003, CAIDAS)

E-Mail (Würzburg): lele.liu [at] uni-wuerzburg.de