Assistant Professor of Data ScienceCity University of Hong Kong (Dongguan)
Efficient and reliable machine learning.
I study model compression, uncertainty, and learning with limited labels across language models, geospatial data, and healthcare.
RA applications open. Join the group.All PhD positions through and including Fall 2027 have been filled.
Efficient & Reliable ML
LowRankArena shows that fewer parameters do not always mean faster inference. By exposing trade-offs between model quality, memory, and latency, it helps researchers judge whether compression delivers useful savings under real deployment constraints.
NeurIPS'26, ED Track, Poster
Co-first author
Geospatial AI
Where palms grow matters for understanding tropical forests and planning conservation. PRISM and PalmDSNet make those spatial patterns measurable from aerial imagery, connecting individual-tree mapping with ecological monitoring across diverse forest landscapes.
IJCAI'25, OralTGRS'25
Co-first and corresponding author
Medical AI
Longer ECG recordings contain temporal context that short-clip models cannot combine on their own. Our work makes that context usable while retaining pretrained backbones, extending the value of existing models without full retraining.
NeurIPS'26, Poster
Co-author
Medical and remote sensing image/signal analysis. Advisors: Prof. Raymond H. Chan and Prof. Jean-Michel Morel.
Ultrasound video analysis and biomedical imaging. Advisor: Prof. Alison Noble.
First Class Honours · Minor in Computer Science.
Delft, Netherlands.