PhD Student, Multimodal Reconstruction of Accelerator Phase Space using Deep Learning | PhD Student (f/m/d) Multimodal Reconstruction of Laser-Electron Accelerator Phase Space...

Job Summary

This PhD position focuses on pioneering multimodal reconstruction techniques for laser-electron accelerator phase space using Physics-Informed Deep Learning. The successful candidate will be responsible for understanding and implementing the physical processes of Coherent Transition Radiation (COTR) in a differentiable form, primarily using pyTorch. A core task involves designing and deploying physics-informed deep neural network architectures to accelerate the self-consistent determination of electron bunch shapes from large measurement campaigns. This role requires strong programming skills (Python, C++, Julia), expertise in deep learning frameworks, and experience working in a Linux/cluster environment. The student will work closely with experimental colleagues, publish scientific results in journals, and present findings at conferences. The position offers a structured PhD program within HZDR’s vibrant, international research community, contributing to critical research in Energy, Health, and Matter, supported by public sector benefits (TVöD-Bund).