Computed tomography (CT) has transformed fields by enabling non-destructive three-dimensional (3D) imaging. However, conventional CT is fundamentally limited by acquisition speed: high-quality 3D reconstruction typically requires thousands of X-ray radiographs acquired over a full specimen rotation, taking minutes to hours. This makes CT unsuitable for fast, transient phenomena such as additive manufacturing, impact, fracture and high-rate deformation.
We are launching an ambitious research programme to overcome this limitation and establish new approaches for X-ray tomography of rapidly evolving processes. Combining advances in X-ray hardware, computational imaging, machine learning and mechanics, the project aims to reconstruct the three-dimensional evolution of materials during dynamic deformation, moving beyond static CT towards true four-dimensional imaging in space and time.
We seek an outstanding Research Associate/Research Assistant to play a leading role in developing the computational foundations of this programme. The successful candidate will develop reconstruction and image-processing methods for highly under sampled, dynamically evolving X-ray datasets, including approaches based on neural radiance fields (NeRFs), neural representations and related state-of-the-art computational imaging techniques.
A central challenge will be to combine X-ray measurements with physical knowledge of deformation processes. The project therefore offers significant scope to develop new physics-informed and data-driven approaches coupling computational imaging with models of material deformation, with considerable freedom to shape the computational research programme.
The computational work will be closely integrated with new laboratory X-ray hardware and high-rate mechanical experiments. Experience or interest in continuum mechanics and computational solid mechanics would therefore be advantageous. Familiarity with finite-element methods, particularly ABAQUS, including coupling ABAQUS to external codes through its Python interface, would also be beneficial.
The project is interdisciplinary and collaborative, involving researchers at the University of Cambridge and industrial partners. Applicants should have a strong background in image processing, inverse problems, machine learning.
Fixed-term: The funds for this post are available for 24 months in the first instance.
Click here to apply online.
Please upload your Curriculum Vitae (CV), list of publications and covering letter. Additional documents cannot be considered. Applications must be submitted by midnight on the closing date.
For questions about the vacancy or application process, please contact Hilde Hambro, Group Administrator: hh463 [at] cam.ac.uk (hh463[at]cam[dot]ac[dot]uk); +44 (0)1223 748243
Please quote reference NM50970 on your application and in any correspondence about this vacancy.