Projects

Design proteins using mask prior-guided denoising diffusion for inverse protein folding

Develop a cost-effective cardiac instability detection tool using multimodal variational autoencoder

Predict molecular properties with geometry-aware line graph transformer pre-training

Advance drug discovery with interpretable bilinear attention network and domain adaptation

Quantify the uncertainty in automatic anatomical landmark localisation

Enable accessible machine learning from multiple data sources for interdisciplinary research

Model both second-order and third-order structures simultaneously for complex networks

Classify autism across multiple sites via site-dependence minimisation & second-order functional connectivity

Construct a feature-importance-aware and robust aggregator for graph convolutional networks

Learn low-dimensional representations of high-dimensional data from their natural tensors