My name is Chenghao Qian, and I am currently a PhD candidate at the University of Leeds working on generative spatial modelling of adverse weather. My research is motivated by two questions: how to model adverse weather for autonomous systems, and how their robustness can be systematically improved. In practice, collecting diverse real-world data in heavy rain, snow, or fog is extremely challenging, and existing datasets often represent weather with coarse labels that miss important physical properties such as particle density, motion dynamics, scattering, and visibility. With CV4DT I plan to delve deeper into geometry- and physics- driven ideas for 3D vision tasks!