Simon(Chengxi) Zeng

I am a Ph.D researcher at Visual Information Lab, University of Bristol, where I am supervised by Dr Alberto Gambaruto and Dr Tilo Burghardt. We are involved in a NIHR-funded project on Swallowing Rehabilitation. My part of work focuses on Medical Image Processing using Deep Learning and Computational Fluids Dynamics using Meshless method.

At Bristol, I also acquired a broad set of skills and subspecialities from my MEng degree in Mechanical Engineering. My undergraduate dissertation is exhibited on Faculty page. Extracurricularily, I served as the President of the Chinese Students & Scholars Association and awarded as President of Excellence by the Embassy of China. And I co-founded an online Education platform - Erdemy Education and it’s been endorsed by Basecamp Enterprise Team from university and granted by UKVI.

Email  /  CV  /  Linkedin  /  Github

My research spans the areas of Computer Vision(Medical Image Processing), Math(Meshless Approximation), Physics(Fluids Dynamics) and Human Physiology( Swallowing Difficulty) .

We propose Video-TransUNet, a deep architecture for instance segmentation in medical CT videos constructed by integrating temporal feature blending into the TransUNet deep learning framework. In particular, our approach amalgamates strong frame representation via a ResNet CNN backbone, multi-frame feature blending via a Temporal Context Module (TCM), non-local attention via a Vision Transformer, and reconstructive capabilities for multiple targets via a UNet-based convolutional-deconvolutional architecture with multiple heads.

We are developing an Emergency drone that delivers medical kits and portable AED that would save lifes in jammed cities and remote areas. This project collborates with Hangzhou Municipal Health Commision and the drone will work with local Emergency Response Unit.

We utilised CNN and other image processing techiniques to non-destructively detect Elastic Transformation on steel/composite materials due to repetitive work load in extreme environments.

We discretise standard Partial Differential Equations(PDEs) using Radial Basis Functions, aiming to assess their performance compared to the analytical results and alternative discretisation methods.

We modelled Bristol city centre area and applied CFD simulation with focus on high wind speed area and presented the advantage of trees to mitigate wind using Porous Media approach.