Qiyuan Xu (Andrew Xu)

I work on machine learning for physical systems, with a focus on computer vision, synthetic data, and sim-to-real learning.

Project Associate, Beam Instrumentation Group, CERN

2024-2026

PhD in Physics, Department of Physics, University of Liverpool

2023–2027

MSc in Artificial Intelligence, School of Informatics, University of Edinburgh

2019-2020

BEng in Electrical and Electronic Engineering, Xi'an Jiaotong-Liverpool University

2013–2017

Publications

Localized Response Basis for Data-Efficient Transverse Beam Distribution Reconstruction Using a Multimode Fiber Relay Qiyuan Xu, H. Zhang, A. Hill, A. Gilardi, F. Roncarolo, G. Trad, M. Bergamaschi, S. Burger, C. Welsch
IPAC 2026

Development of radiation-tolerant beam imaging via multimode fiber and synthetic data-driven machine learning Qiyuan Xu, H. Zhang, G. Trad, A. Hill , F. Roncarolo, C. Welsch
Physical Review Accelerators and Beams

Evaluating Machine Learning Models for Multimode-Fiber-Based Transverse Beam Profile Reconstruction Qiyuan Xu, A. Hill, H. Zhang, F. Roncarolo, G. Trad, S. Burger, W. Farabolini, P. Korysko, D. Metin, C. Welsch
IBIC 2025

Estimation of Beam Transverse Parameters Through a Multimode Fiber using Deep Learning Qiyuan Xu, A. Hill, H. Zhang, F. Roncarolo, G. Trad, C. P. Welsch
IBIC 2024

AuleLab – decoupled from OptiBeam, optimized for multi hardware control and sync, under development...

XFlow – minimal, modular machine-learning framework for scientific workflows

OptiBeam – my PhD experiment hardware control library and automation framework

IBIC 2026 invited talk

NVIDIA & Dell AI Discovery Day poster prize

Best Young Researcher Poster at MaLAPA'25

Best poster prize at Cockcroft Institute conference