United by excellence and a common passion for knowledge, I am attracted to the Hong Kong Laureate Forum as it presents a unique opportunity for world-leading and young researchers across the globe to share and learn. I believe this mixture of diverse backgrounds creates the necessary environment for stimulating new ideas and collaborations needed to solve our world's most pressing problems. I particularly look forward to talks by Shaw Laureates Dr Mori and Dr Walter, as their ground-breaking work on the Unfolded Protein Response has surprising implications in Parkinson's disease, whose involvement of the mesolimbic reward network is closely related to my research in music neuroscience.

Although I have a background in mathematics and computational neuroscience, as well as a doctorate in psychology, I now work as an emotional AI researcher in the music industry. For me, this unique combination of skills and experiences have proven valuable in my research. I am keen to contribute to this forum by sharing my work and offering insights on how academic research can be translated to applications in everyday life. I also look forward to meeting other like-minded scientists who share the same passion for knowledge via innovation and interdisciplinary research.

Despite serving no direct biological purpose, music is often regarded as one of life’s greatest pleasures. My research is about understanding why people enjoy music from a neural-cognitive perspective and using these insights to develop next-generation music recommendation systems with artificial intelligence.

A key research theme is the encoding of musical pleasure. I use neuroimaging methods such as functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) to study how music is processed and manifested as a reward stimulus in the human brain. For example, our team showed that changes in musical expectancy is related to pleasure and activity in the mesolimbic reward network. We are now currently developing EEG systems that aims to decode music preference in a single-trial, real-time setting.

A complementary research focus is the development of machine learning methods and models. One example is our work on automatic voxel selection in decoding from fMRI data that capitalises on whole-brain data to maximise performance. Another example is time-warped representational similarity analyses, which enables the correlation between multimodal signals (e.g., neural, behaviour, and audio features) with variable duration. Currently, we are developing transfer-learning techniques that enhance brain decoding performance from model pretraining using large-scale datasets.