Summary
Guoxuan Xia is a PhD student at Imperial College London with eight years of hands-on experience bridging research and machine learning engineering, focusing on efficient deep learning and uncertainty estimation. He has interned at leading industry labs including Huawei, MediaTek Research, Arm, and contributed to practical advances in multimodal generative models and computationally efficient diffusion sampling. His work blends model compression, distillation and pruning for transformer deployment with novel noising schedules and fast ODE samplers for image generation, showing a strong orientation toward deployable research. Past projects include camera-parameter recovery for historical illustrations and drone-based validation, demonstrating creative application of computer vision and optimisation. Based in London with an MEng from Cambridge, he brings both rigorous academic training and product-aware experimentation to ML problems. He often focuses on the less obvious trade-offs—like uncertainty-aware efficiency—that make models both reliable and practical in real-world systems.
8 years of coding experience
Master of Engineering - MEng, ENGINEERING, Honours with Distinction, Master of Engineering - MEng, ENGINEERING, Honours with Distinction at University of Cambridge
A-levels, A* Maths A* Further Maths A* Physics A* Chemistry, A-levels, A* Maths A* Further Maths A* Physics A* Chemistry at Royal Grammar School Newcastle
Chinese